is it time for automatic voter registration in the uk? · 1.3 there has been cross party support...

Is it time for Automatic Voter Registration in the UK?

Toby S. James & Paul BernalUniversity of East Anglia

This report was funded by the UK Democracy Fund.

The UK Democracy Fund, a Joseph Rowntree Reform Trust Ltd initiative, funded the work reported in this publication. The material presented here represents the views of the author(s), not necessarily those of JRRT or other UK Democracy Fund contributors.

Toby S. James is Professor of Politics and Public Policy at the University of East Anglia.

Paul Bernal is Associate Professor in Information Technology, Intellectual Property and Media Law at the University of East Anglia.

The 2014 Research Excellence Framework (REF 2014) rated 70% of UEA’s research in Politics and International Studies and Law as 4* (world leading) or 3* (internationally excellent), as well as ranking UEA eighth in the UK for its research impact in Philosophy (Times Higher REF 2014 Analysis) and joint fifth in the UK for the quality of its research in Area Studies (Times Higher REF 2014 Analysis).

Toby. S. James and Paul Bernal (2020) Is it time for Automatic Voter Registration in the UK? Joseph Rowntree Reform Trust: York.

Foreword | 3

Foreword

‘Voting is an essential part of the democratic process. It’s vitally important that the electoral process is fit for purpose for the twenty-first century, and that voting is accessible and convenient as could be to all citizens. This research-led report helpfully sets out the arguments and issues to be weighed up as policy makers and parliamentarians consider whether to introduce automatic voter registration.’

Andrew Bowie MP, Chair of the APPG on Democratic Participation

4 | Executive Summary

This report draws from academic literature, sources including interviews with stakeholders, comparative international experiences and Freedom of Information requests to outline and assess the arguments. It finds that:

• “Automatic voter registration” (AVR) is the direct enrolment of citizens onto the electoral register by public officials, without the need for pro-active action by citizens. AVR can be contrasted with the current system found in the UK, in which citizens are responsible for registering themselves ahead of the deadline for election day. It can also be distinguished from ‘assisted’ voter registration, which involves citizens being asked to register to vote when accessing other government services. AVR is therefore a broad principle and umbrella-term that could in practice involve a range of different approaches.

• Some form of direct enrolment is the norm in democracies around the world because many countries are able to use population registers to directly enrol citizens. Many parts of democracies in the Anglosphere who have not historically had direct enrolment such as Australia, Canada and the USA, however, have recently implemented it for specific under-registered groups such as young people or when accessing specific government services such as driver licence agencies.

• Electoral registration has seen some major modernisation in the UK, following nearly a century of continuity. Recent reforms have included online voter registration, individual electoral registration and annual canvass reform. AVR has been proposed in Wales following further devolved powers.

Executive Summary

Automatic voter registration (AVR) has been increasingly proposed in the UK by parliamentary committees, campaigners, politicians and academics. There are millions of citizens incorrectly registered or missing from the electoral register entirely. Is it therefore time for AVR to boost voter participation? Or are there practical or philosophical problems involved in registering people automatically?

Executive Summary | 5

• There are many underlying problems with the status quo such as millions of eligible citizens being incorrectly registered or missing from the registers entirely, major strains on the system during a last-minute registration rush ahead of election day and resource problems for electoral officials.

• The main philosophical argument raised against AVR is usually that it should be an individual responsibility to register to vote. There are also concerns that AVR might require a civil population register.

• The arguments in favour of AVR are primarily that voter registration rates are in decline and citizens often think that they are already registered. AVR could therefore prevent some citizens being unable to vote on election day or would improve their overall experience. Political equality is a founding principle of democracy and policies are therefore needed to ensure a level playing field on election day. AVR could boost voter registration rates amongst under-registered groups to create this more level playing field.

• There are multiple options for implementing AVR. These include:

1. A new centralised civil population register. This would be the most direct pathway to AVR, but concerns about civil liberties may make this approach unpopular and it would have much broader consequences.

2. The expansion of the DWP Customer Information System to act as a single national electoral register.

3. Localised data-mining. AVR could be introduced by making datasets available to local electoral registration officers to identify and register missing voters. Past experience has shown that this approach is not necessarily cost effective and it could be very unevenly implemented.

4. Direct enrolment for specific groups only, such as 16 year olds.

5. Assisted voter registration. This would involve citizens being prompted to register to vote when accessing a range of other government services such as applying for a driving licence.

This report outlines what AVR is, and considers the arguments for and against

its adoption in the UK. It makes 28 recommendations to modernise the

UK’s electoral machinery for an inclusive twenty-first century democracy.

1 sub

• The report provides data on the current frequency of citizens’ transactions with some key government services. It suggests that options 4 and 5 (direct enrolment for specific groups and assisted voter registration) could be the most cost-effective methods that would lead to considerable improvements in the completeness and accuracy of the register.

• Directly registering citizens shortly before their 16th birthday when they are issued their National Insurance Number would add 700,000 citizens to the roll each year with minimal administrative effort. These citizens could then be provided accompanying civic education lessons while in school.

• Providing citizens with an option to register to vote/update their registration details when accessing other government services would enable millions of citizens to register more easily. For example:

- 6.5 million per year could register when applying for a passport

- 4 million people could register when they update their driving licence address with the DVLA

- 2 million a year could register when applying for Universal Credit

- 2.5 million students could be registered through annual student enrolment

- 800,000 could register when they apply for child benefit for the first time

- 500,000 could register when they provide the Student Loans Company with a new address

- 450,000 could register when they apply for disability benefits

• Accompanying reforms are also recommended to overcome issues with privacy and data security. New direct enrolments should not be added to the open/edited electoral register (which can be bought by anyone) and this register should be abolished. Anonymous registration should be extended to at least five years and the procedures to register anonymously should be relaxed to prevent vulnerable adults being at risk of harm. There should be tightened restrictions on the use of the electoral register for non-electoral purposes.

• Overall, this report makes 28 recommendations to modernise the UK’s electoral machinery for an inclusive twenty-first century democracy.

6 | Executive Summary

Contents

Introduction 9

Towards AVR? A Policy History 12

How Electoral Registration Works in the UK 14

Nine Problems with the Status Quo 21

The Overseas Experience 28

Philosophical Choices and Issues 34

Options for Implementing AVR 41

Finding the ‘Life Moments’: New National Data Sources for Direct Enrolment 44

Conclusions and Recommendations 55

Appendix A: Survey Questions 58

Appendix B: Respondents 59

Appendix C: Government Transactions with Citizens 60

Bibliography 61

1 sub

“Automatic voter registration” (AVR) is the direct enrolment of citizens onto the electoral register by public officials, without the need for pro-active action by citizens.

1 The term is used interchangeably with ‘direct enrolment’ and ‘automatic voter registration’ in this report – to which it is more commonly referred to in some countries including Australia. Introduction | 9

What is ‘Automatic voter registration’?

1.1 “Automatic voter registration” (AVR) is the direct enrolment of citizens onto the electoral register by public officials, without the need for pro-active action by citizens. AVR can be contrasted with the current system found in the UK, in which citizens are responsible for registering themselves ahead of the deadline for election day.1 AVR can also be contrasted with assisted voter registration – which is defined in this report as a range of interventions to prompt citizens to register to vote when they interact with other government services.

1.2 AVR has been described as a potential gamechanger for democratic politics. In an age where many citizens do not vote, often because of confusion about voter registration, it has been positioned as a critical structural problem that could have the most impact on the number of citizens registered. While voter registration does not automatically translate into voter participation, there is evidence that citizens are often unable to vote because they think that automatic registration already exists and are turned away on election day. Being on the electoral roll enables citizens to be contacted by parties and pressure groups encouraging them to participate. It is therefore seemingly an important step for realising an inclusive electoral process.

1.3 There has been cross party support for AVR in the UK from the Political and Constitutional Reform Committee in its 2014 report on Voter Engagement (Select Committe on Political and Constitutional Reform 2014), and the All-Party Parliamentary Group on Democratic Participation’s report on the Missing Millions (Bite the Ballot, James, and ClearView Research 2016). The Electoral

Commission called for automatic voter registration processes in 2017 (Electoral Commission 2017b).

1.4 There have also been important concerns raised about the potential impact of automatic registration on the long-term accuracy of the electoral register, on privacy, on data protection and security. These issues therefore need detailed consideration if the UK is to continue to modernise the electoral registration process utilising technology and automation – as it increasingly has done since the turn of the century.

The Purpose of this Report

1.5 The purpose of this report is to understand the potential trade-offs and choices available to policy makers, identify safeguards that could address concerns and examine how the UK could introduce various forms of assisted, partially automated or fully automatic registration. The research questions that it therefore seeks to address are as follows:

- How does the system currently work? - What are the problems with the existing

system of electoral registration? - What alternative systems have been used

overseas? How effective have they proven? - What challenges might there be for

implementing AVR? - How could AVR work in the UK?

1.6 The Electoral Commission has recently published a feasibility study, which makes considerable inroads into this topic. There remain several broader legal and organisational issues to be considered. The Electoral Commission study focussed on whether AVR was feasible in law, but not whether the broader legal framework

1 | Introduction

2 Based on James (2020, 59-86) https://tandfbis.s3-us-west-2.amazonaws.com/rt-files/docs/Open+Access+Chapters/9781138682412_oachapter04.pdf10 | Introduction

sufficiently protected the rights of citizens. This report also draws from a wider set of academic and comparative research about different voter registration systems.

What does good electoral registration look like?

1.7 Before we begin, it is worth considering what a modern electoral registration system should look like. Electoral registration should aim to have2:

• Complete registers – everyone who is eligible to vote should be included. This is important to ensure that no one is effectively disenfranchised through administrative issues.

• Accurate registers – all names on the electoral register should be accurate and eligible electors. This is important in the prevention of electoral fraud.

• Cost-efficiency – the register must be delivered in a way that is cost efficient. This is important because public resources are always finite and excess funds could be used elsewhere.

• Investment sufficiency – sufficient funding should be made available to compile the electoral register. This is important so that no ‘corners are cut’ in the delivery of electoral registration.

• Citizen convenience – minimal effort should be exerted by a citizen to register to vote. This is important because it is an attribute of the quality of service but could also affect whether citizens register.

• Equity of outcome across society – registration and accuracy rates should be consistent across the population so that no one group is disadvantaged or advantaged. This is important because different groups could receive unequal representation and election results might be affected by a ‘turnout gap.’

• Robust staff capacity – there should not be

a disproportionate workload burden on staff who should be motivated, have the required skills and ability, and work in good working conditions. This is important for the delivery of an election without risks of errors.

• Robust service functionality – potential vulnerabilities should be identified and addressed that could affect the deliverability of the election and undermine voter confidence. This allows problems to be prevented in advance.

1.8 In addition, there are further issues relating to data transparency/citizen rights that have not always been given due consideration. People should be aware of how and when their data will be added to the system, for example when their interactions with government websites might trigger an entry

or change to the record, and have opportunities to access and correct that data, as well as knowledge and understanding of all the functions to which the registers will be used.

1.9 It is also important that citizens are aware of the external use of that data. As this report will show, the electoral register is not just used for the purpose of running elections. It is also used for commercial activity and the prevention of other government objectives such as anti-fraud measures. These issues with

transparency and citizen rights have been less of a feature of the debate on AVR to date. This report aims to bring these issues to the fore.

The need for proactive voter registration modernisation

1.10 The world does not stand still. There have been major changes in the nature of society since the voter registration system was first developed in the context of the First World War. These changes include a much more mobile population, greater diversity and changed expectations from citizens about how they think governmental bodies work –

The world does not stand still. There have been major changes

in the nature of society since the voter

registration system was first developed in the context of the First

World War.

Introduction | 11

and how they should work. All government services need to keep track with these developments or there will be a decline in their effectiveness and performance.

1.11 Well run elections also require proactive steps to address gaps in participation and under-registration. Drifts towards increasing under-registration don’t affect everyone evenly. As this report shows, it is increasingly the young that are missing from the electoral register. This can have profound consequences for electoral boundaries, representation and voter participation.

1.12 Electoral modernisation is therefore an important part of any government’s agenda. It has been so in recent years, with reforms such as the move to continuous registration, individual electoral registration and the modernisation of the canvassing process. This report, however, aims to consider whether direct enrolment should be part of the next steps of this modernisation.

Report Structure and Approach

1.13 This report involved: - Desk research of policy documents and

academic research on voter registration in the UK and overseas

- Interviews with UK-based stakeholders in the electoral community, alongside interviews with overseas organisations who have already implemented AVR to learn lessons from their experiences

- Data collection on international electoral registration systems based on election observation reports and other sources.

- An open web-survey through which stakeholders could provide further input.

- A site visit to an electoral registration office to see the current system in action.

- Freedom of Information requests to several government departments and governmental agencies to ascertain the volume of applications for/transactions with a variety of government services.

1.14 The report begins with a brief policy history of how we got to where we are today in Chapter 2, a stone’s throw away from AVR. Chapter 3 then outlines how the electoral registration currently functions, including the recent move to automatic re-registration.

Nine fundamental problems are set out in Chapter 4, which the move to AVR could, according to its supporters, address in some respect. Chapter 5 puts the UK into international perspective by looking at how voter registration is run overseas and how AVR has been introduced in other countries. Chapter 6 considers some of the ethical, legal and practical arguments and challenges that AVR would bring. Chapter 7 considers options for how AVR could work, if it was to be introduced. Chapter 8 expands on this by providing information on the ‘moments’ where voter registration could be integrated into a citizen’s life in the future. The conclusions and recommendations are then summarised out in Chapter 9.

12 | Towards AVR? A Policy History

2.1 The UK’s electoral registration system, like much of the nation’s infrastructure, has Victorian origins. Electoral registers were used for the first time after 1832, where the duty was given to the parish officers who oversaw the poor laws. They invited citizens to claim to be on the register in June each year. Claimants would be excluded from the register if they had not paid their taxes. A list was openly published in July and provision made for objections to be made about any name on the register. In the case of a dispute claimants and objectors would be required to attend court. In practice, much of the work to register voters was done by political parties, who also worked to often object to the entries brought in by their opponents (James 2012, 125-68; O’Leary 1962; Seymour 1915).

2.2 The First World War provided the next major trigger for change. The war led to major human displacement which meant that earlier registers were outdated. Meanwhile, the expansion of the franchise meant that many more people needed to be added to the register. The Representation of the People Act (RPA) 1918 established the new system in which there would be regular door-to-door canvassing of properties to ensure that the register was as accurate and complete as possible. The RPA1918 made provision for two registers per year to be made, but this was reduced to one eight years later to save money as part of the Economy (Miscellaneous Provisions) Act 1926.

2.3 Very little change occurred until the start of the twenty-first century and the election of New Labour who undertook a broader range of changes to how the electoral machinery functioned. The Representation of the People Act 2000 introduced a move from a periodic system to a continuous

system. There were also some significant moves towards the introduction of individual electoral registration (IER). Britain had long had a system of household voter registration whereby one ‘head of household’ could complete a registration form on behalf of all citizens living in a household. IER was introduced into Northern Ireland in 2002 and legislation was being prepared to bring it into force across the remainder of the UK by the New Labour government.

2.4 The Coalition government fast-tracked the move to IER. The Electoral Registration and Administration Act 2013 stated that every single citizen would be required to register individually. They would also be required to provide their date of birth and a National Insurance number as personal identifiers. The system had already been introduced to Northern Ireland, but it would now run across the UK. The Act also introduced a move to online voter registration and the checking of new entries against government databases. The legislation also ended plans, developed by the Labour government, to develop a single electoral register as part of the CORE project. Mark Harper, Minister for Political and Constitutional Reform, said that ‘the costs of building and running the database are disproportionate to its potential benefits’, quoting an estimated £11.4 million implementation cost, and £2.7 million per annum to run (UK Government 2011).

2.5 IER became a reality across the rest of the UK in 2014 after which all citizens had to register under the new system. Names were initially carried over from the old registers until December 2015, after which citizens who had not registered individually were removed. Concerns were raised about whether the completeness of the electoral register had been affected by the change.

2 | Towards AVR? A Policy History

Towards AVR? A Policy History | 13

Academic studies predicted that students and young people were less likely to be registered as a result (James 2014). The Political and Constitutional Reform Committee, notably in its 2014 report on Voter Engagement (Select Committe on Political and Constitutional Reform 2014) therefore voiced concerns. This was followed by the All-Party Parliamentary Group on Democratic Participation’s report on the Missing Millions (Bite the Ballot, James, and ClearView Research 2016). Data on the eventual trends are summarised in Chapter 4.

2.6 Reforms followed to address problems with the completeness of the register and deal with some of the other side effects of the new process. The Higher Education Research Act 2017 introduced a requirement for universities to encourage their students to register to vote.3 Further modernisation followed in 2019. The Minister presented the Representation of the People Act (Annual Canvass) (Amendment) Regulations 2019, which brought an important move towards automatic re-registration by reforming the canvassing process from 2020 onwards.4

2.7 The Electoral Commission gave support to connecting voter registration more closely with other public services, in its report on the 2017 general election (Electoral Commission 2017b). It then published some feasibility studies in 2019 to examine whether EROs could be given access to other forms of data. It reported that it would be possible from a ‘technical and operational perspective and could be implemented without radically altering the structure of the electoral registration system in the UK’ (Electoral Commission 2019d, 1).

2.8 Recent legislation has given Scotland and Wales greater power to manage the register for local government – the register which is also used for their respective Parliaments.5 This has led to discussions about whether reforms could be introduced in these jurisdictions without UK-wide reform. Wales and Scotland also passed legislation to lower the franchise to 16 for these elections.6

2.9 The Welsh government has taken forward recommendations from the Voter Engagement report to introduce the Local Government and Elections (Wales) Bill, currently before the Senedd at the time of

writing, which would introduce AVR as it ‘provides for a power for EROs to add an individual to the electoral register without the need for them to apply.’ It would also establish a framework to facilitate a single electoral register for Wales.7

2.10 The Scottish government introduced the

Scottish Elections (Reform) Bill in September 2019 to alter the role and remuneration of Returning Officers,8 but more ambitious reforms for electoral registration were not taken forward, as the Standards, Procedures and Public Appointments Committee noted.9

Summary2.11 There has been major public policy concern

about whether the UK electoral registration system is fit for purpose and greater interest in AVR. There are now new devolved powers to Scotland and Wales to manage their electoral registers for local government and devolved elections. This is therefore a critical juncture in which there is an historic opportunity to further modernise the UK’s electoral registration, and opportunities to explore further innovation in Scotland, Wales and Northern Ireland.

3 http://data.parliament.uk/writtenevidence/committeeevidence.svc/evidencedocument/electoral-registration-and-administration-act-2013-committee/electoral-registration-and-

administration-act-2013/written/106497.html4 http://www.legislation.gov.uk/uksi/2019/1451/pdfs/uksiem_20191451_en.pdf

5 The Wales Act 2017 and the Scotland Act 20166 Senedd and Elections (Wales) Act 2020.

7 https://www.assembly.wales/laid%20documents/pri-ld12877-em/pri-ld12877-em-e.pdf p. 14-7.

8 https://www.parliament.scot/S5_Bills/Scottish%20Elections%20%20(Reform)%20Bill/SPBill53PMS052019.pdf

9 https://sp-bpr-en-prod-cdnep.azureedge.net/published/SPPA/2020/1/14/Stage-1-Report-on-the-Scottish-Elections--Reform--Bill/SPPAS052020R01.pdf

14 | How Electoral Registration Works in the UK

Who is responsible for running electoral registration in the UK? How does the electoral registration system currently work? This chapter answers both of these questions in turn.

Electoral Registration Today: Who does what?

3.1 Parliament in Westminster alone has the power to set legislation governing the UK Parliamentary register although Ministers can use secondary legislation for some changes. Legislative competence for the local government register in Scotland was devolved to the Scottish Parliament by the Scotland Act 2016. Legislative competence for the local register in Wales was devolved to the National Assembly for Wales by the Wales Act 2017. Scottish and Welsh Ministers can both make changes to the respective processes as exercisable by a Minister in Westminster, provided that they were within the powers of devolved competence.

3.2 There is a very decentralised system for registering voters and maintaining the register. Responsibility for maintaining the electoral register resides with the Electoral Registration Officer (ERO). They are an employee of local government, but are independent of local government for their statutory responsibilities relating to the electoral register. They are instead accountable to the courts system as an independent statutory officer and can be prosecuted for being in breach of their duties. EROs are appointed by the appropriate local council(s).

3.3 EROs are supported by an administrative team in each local authority which tends to consist of a Democratic Services Officer who provides strategic level management, an Electoral Services Manager who manages the day-to-day running of the elections

department and makes preparations for the poll and an administrative support team. In most local authorities this team tends to be very small, but might be boosted by several short-term appointments during the immediate election period.

3.4 In Scotland, EROs are appointed by a combination of local authorities. The ERO also serves as the Assessor in the Joint Valuation Boards (JVBs), who have responsibility for the administration of local taxation and to compile the electoral register. A Chief Electoral Officer acts as both the Returning Officer (RO) and Electoral Registration Officer in Northern Ireland.

3.5 The Electoral Commission was established in 2000. It did not originally have any direct role in the registration of electors, however, until it was granted powers to set performance standards for EROs, as well as ROs. Its role is otherwise in the provision of guidance to electoral officials and the creation of pooled resources. There are also many organisations such as the Electoral Management Body for Scotland, Association of Electoral Administrators, SOLACE, SOLAR and Scottish Assessors Association who play a role in the development of training and guidance for electoral officials, but do not play a direct role in the management of the electoral register either.

3.6 Like Victorian pipework, or early underground train systems, the system for electoral registration has therefore been one of incremental change with systems layered on top of old infrastructure. It has become increasingly complex over time, partly as a result of devolution.


1 sub

Like Victorian pipework, or earlyunderground train systems, the system for

electoral registration has therefore been one of incremental change with systems

layered on top of old infrastructure. It has become increasingly complex over time,

partly as a result of devolution

10 In practice, few, if any EROs issue civil penalties for non-registration so it is effectively unenforced.16 | How Electoral Registration Works in the UK

How the Electoral Registration Works

How voters apply to register3.7 Registration is compulsory in the UK and

citizens can be fined for not registering.10 Since 2014 it has been an individual’s own responsibility to register and they cannot rely on a parent, carer or housemate to register them on their behalf.

3.8 There are two main pathways through which citizens can apply to register to vote.

• An online application via the UK government’s website (Figure 1.1).

• The annual canvass. This is a yearly check on the completeness and accuracy of the electoral register at each property. As paragraph 3.28 explains, this will be revised for the 2020 canvass, but involved two-stage process after the introduction of IER. First, Household enquiry forms (HEFs) were sent to all households between July and November. These detailed existing electors known to be registered at the address. Households were then asked to confirm the details and notify the ERO of any changes. Second, Invitation to Register (IRV) forms were sent those who identified as missing from the register, following the return of a HEF. Only once an IRV form has been completed would an application be processed.

Figure 1.1: Online voter registration form

3.9 In both cases citizens are asked to provide their: • full name • date of birth • current address • previous address where they have ceased

to reside in the last 12 months

• National Insurance number (or information about why they are not able to do so)

• Nationality • Whether they wish to be excluded from the

edited/open register • A declaration that the contents of the

application is true • The date of the application

The New Annual Canvass System

Automatic re-registration goes live in the UK in 2020.

Citizens who can be matched against national and local data sources will no longer need to reply to the canvass if their details are correct (described as the ‘Route 1’ pathway).

EROs will continue to canvass those who can’t be matched via data sources (‘Route 2’ pathway).

Those who live in specific property types can be re-registered by a ‘gatekeeper’ such as a building manager or care homeowner (‘Route 3’ pathway).

Processing applications

3.10 Each application is run through the Cabinet Office’s Individual Electoral Registration Digital Service (IERDS) to verify the applicant. This attempts to match the applicant’s National Insurance number (NINo), date of birth and name against the DWP Customer Information System (CIS). The CIS contains a record for all individuals who have registered and been issued with a National Insurance number which is needed to work or claim benefits (DWP 2018).

3.11 EROs are provided with information from the IERDS as to whether there is a ‘pass’ or ‘fail’ in the matching process, which they then use to assess whether they should be added to the electoral register. EROs are also allowed to use locally held data to make a decision (see: paragraph 3.18). The digital matching process does not check nationality so is not a guarantee of eligibility.

11 Such analysis would also be difficult because EROs are exempt from Freedom of Information. How Electoral Registration Works in the UK | 17

Removing names

3.12 Names remain on the electoral register indefinitely, unless the ERO assesses that an individual has ceased to be resident, is not eligible or that the application was made by a third party. In practice, this means that names are removed when the ERO is informed either by receipt of an HEF form or new application that they have moved. They are also encouraged to explore local data sources, however. Registered electors can also raise formal objections to query the status of other electors. Names can only be removed upon receipt of two pieces of evidence – a response to the HEF will only consist of one these sources. If there is no response to a canvass then EROs can issue a penalty (Hone 2014). The authors are not aware of any systematic collection of data on the number of fines issued by EROs. 11 In practice, however, interview evidence conducted by one of the authors with EROs suggested that fines are very rarely issued (James 2014). Names will remain on the register in the event of a non-return (Electoral Commission 2019c, 36-40).

Current attempts to boost voter registration

3.13 Voter outreach activity to encourage citizens to register to vote has been initiated by the Electoral Commission at election time as the deadlines approach. This is often in collaboration with other actors such as the UK government and NUS, and national TV programmes such as Gogglebox in 2015 (Electoral Commission 2015, 34-5).

3.14 There is a legislative requirement for EROs, under ‘Section 9A of the Representation of the People Act 1983 to take all necessary steps to comply with the duty to maintain the electoral register and to ensure, as far as is reasonably practicable, that all those eligible – and no others – are registered in it’ (Electoral Commission 2019c, 4).

3.15 The Electoral Commission also requires EROs

to develop a public engagement strategy for voter registration activity in order to meet one of their performance standards (Electoral Commission 2016). This strategy should include:

• identifying and using data sources to highlight potential sources of electors who

are less likely to be registered to vote • identifying internal and external partners to

encourage higher voter registration rates • ward level data analysis of the area,

identifying priority areas • Specifying the direct contact, media and

advertising routes for those communities. • An evaluation of the success (Electoral

Commission 2016, 3-4)

3.16 Following a slimming down of the performance standards system, the Electoral Commission only select a small sample of EROs that they think are ‘at risk’ of not meeting the standards for assessment (Electoral Commission 2018). There is therefore no regularly published picture of the extent to which all EROs undertake the voter outreach activity. One academic study, however, reported that 16 per cent of EROs had no such strategy when asked as part of a survey in 2016. Falls in budgets were associated with the absence of a strategy (James and Jervier 2017, 4-5). The same study also reported that school visits were undertaken by some local authorities to encourage voter registration. These were more likely to take place in London boroughs, and less likely in rural areas. Austerity had also affected the likelihood that these would be undertaken (James and Jervier 2017, 5).

Local data mining

3.17 Although the move to AVR is thought to be a dramatic game-changer, EROs already undertake data-mining activities to improve the quality of the electoral register. Electoral Commission guidance states that EROs can demand access to locally held information by local authorities to improve the electoral register and that such request cannot be declined:

‘Paragraph 1(5) of Schedule 2 to the Representation of the People Act 1983 provides that where the ERO requests to inspect and/or take copies of the records specified in paragraph 2.42, a statutory or other restriction, including the GDPR, cannot be used to refuse disclosure of those records.’ (Electoral Commission 2019c, 15)


3.18 These data sources would include, but are not limited to Council Tax records, Council Tax reductions, housing benefits, the Register of households in multiple occupation (HMOs), records held by the registrars of births, deaths and marriages, lists of residential and care homes / shelters / hostels, lists of disabled people receiving council assistance; land registry/registers of Scotland; planning and building control; list of new British citizens held by the registrar; and local authority education data (Electoral Commission 2019c, 15-7).

3.19 The extent to which they do this, however, varies considerably across local authorities with some very proactive, and others much less so.

Multiple registers

3.20 The concept of the UK electoral register is a misnomer. In reality, there is no single electoral register, but a patchwork of registers. There are different electoral registers for each geographical area covered by the ERO. There are also different electoral registers within the geographical areas covered by EROs because there are different franchises for different election types. The Parliamentary register contains all citizens eligible for general elections (which excludes EU citizens). Local government registers contain all electors eligible for local, Scottish and Welsh government elections (which includes EU citizens).

Software

3.21 Data is stored and managed by EROs using a variety of different software packages. As of 2017, there were four major private sector electoral management systems, and seven in-house packages developed by local authorities to manage 373 registers (Electoral Commission 2017a, 6).

Public availability of registers

3.22 The electoral register has never been a secret document. In fact, the availability of the electoral register has been an important part of the transparency and integrity of

the system. The register is open for public inspection at any local authority, under supervision. This allows citizens and political parties to check for erroneous and missing entries. Past registers are also kept in libraries and archives for historical research. There are strict limits that prevent extracts being digitally recorded, scanned or photographed (Electoral Commission 2019c, 169-85).

3.23 Data from the electoral register is also electronically shared in a number of ways for electoral purposes, preventing crime, credit reference checking and jury summoning (Electoral Commission 2019e). EROs will share data with:

• Other EROs when electoral events have electoral constituencies that cross the local authority borders. This is necessary to create constituencies and issue poll cards.

• The Boundary Review Commissions for the purpose of the boundary reviews. They will also provide it to the Electoral Commission.

• Elected representatives, candidates, their party agents, political parties, third parties that are registered with the Electoral Commission.

• Governmental organisations such as the all police forces in Great Britain, Security Service and the National Crime Agency.

• Credit reference agencies for the purposes of credit referencing and anti-money laundering.

3.24 The data tends to be exported from local software systems as a .csv file. This is then either emailed or transferred via data transfer services. Upon receipt of the data, EROs at a neighbouring authority will import it into their electoral management software system.

3.25 It is possible to register anonymously, subject to strict criteria (see paragraphs 4.26-4.29).

3.26 EROs are required to produce an ‘edited register’ or ‘open register’ which is available for purchase by any individual, company or organisation. Citizens have the opportunity to opt-out of this when they register to vote. In Scotland, any citizen under 16 years old is automatically opted out of the electoral register.12 There are no restrictions on the sale of the edited version of the register or on the uses that can be made of it.

How Electoral Registration Works in the UK | 19

Existing Data Protection Provisions

3.27 There are some legal safeguards in place to maintain data protection. The Electoral Commission guidance explains that EROs are considered as the ‘data controllers’ with a statutory duty to process personal information in the task of maintaining the electoral register. They are therefore expected to demonstrate that they ‘comply with the principles of processing personal data, ensuring that it is processed lawfully, fairly and in a transparent manner’. EROs do not need to appoint data protection officers (DPO) because they are not a ‘public authority’, but are advised by the Electoral Commission to liaise with the DPO of their council (Electoral Commission 2019b, 7).

Canvass reform and automatic Re-registration

3.28 In November 2019, Parliament passed the Representation of the People Act (Annual Canvass) (Amendment) Regulations 2019. This was a statutory instrument which received very little media interest coming immediately before the general election, but represents an important evolution of the electoral registration system, which brought an important move towards automatic re-registration by reforming the canvassing process from 2020 onwards.13 The changes followed from a government consultation on annual canvass reform, which involved a collaboration with the Scottish and Welsh governments and was launched in October 2018.14

3.29 From 2020 all citizens will not be canvassed in the same way. The canvass process will begin with a new ‘data discernment step’ in which properties on the register will be matched against national and local data sources. Where a match has been made, EROs will be required to send a written letter to each property asking for notification of changes, but citizens will simply be re-registered unless there is a response.

3.30 Where national and/or local sources do not match against the electoral register, the residents of the property will be contacted asking them to respond to the canvass. EROs are required to chase non-responses, and there must be at least three attempts to contact the property, at least two of those

must have involved visits to the property.

3.31 A specific path is set out for unique properties such as registered care homes, student accommodation, HMOs and hostels. In these cases, a ‘responsible person’ can be approached for a list of the names of the residents. If no response is received within a reasonable time period, then a full canvass of that property must be conducted.

3.32 The requirements for removal from the register remain the same (in that two pieces of evidence are usually required). Names cannot therefore be deleted just by virtue of a non-response to the canvass.

Summary3.33 Electoral registration has already become

highly digitalised. It already involves EROs having access to multiple databases and using data mining skills to ensure that the electoral register is as complete and accurate as possible – as well as knowledge of how laws such as the GDPR apply to this data access and data mining. The potential move to AVR has therefore been eased by these recent reforms.

12 Section 13 Scottish Elections (Reduction of Voting Age) Act 201513 ://www.legislation.gov.uk/uksi/2019/1451/pdfs/uksiem_20191451_en.pdf

14 https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/833308/Reform-of-the-Annual-Canvass-Statement-of-Policy.pdf

1 sub

Electoral registers should include all citizens who are eligible to vote in an election. This is important, not only so that they can vote on election day, but because the electoral register is used for other civic purposes such as jury service and the construction of electoral boundaries.

Nine Problems with the Status Quo | 21

15 https://www.electoralcommission.org.uk/who-we-are-and-what-we-do/our-views-and-research/our-research/accuracy-and-completeness-electoral-registers/2019-report-2018-electoral-registers-great-britain/completeness-great-britain

This chapter sketches out nine problems to the existing voter registration system that AVR has often been cited as having the potential to fix.

The Missing Millions

4.1 Electoral registers should include all citizens who are eligible to vote in an election. This is important, not only so that they can vote on election day, but because the electoral register is used for other civic purposes such as jury service and the construction of electoral boundaries. The completeness of the electoral register is defined by the Electoral Commission as the extent to which every person who is entitled to have an entry on an electoral register is registered’ (Electoral Commission 2019a).

4.2 The completeness of the register is not the only issue. Electoral registers should also be accurate so that every name entered on the register is a valid, eligible citizen with correct

registration details. A citizen should usually only appear once on the register (although there are exceptions, such as students who can register in two locations, but only vote once).

4.3 Measuring the completeness of the electoral register is beset with measurement problems (Wilks-Heeg 2012, 60-72). Estimates from the Electoral Commission’s recent studies provide the most reliable measures of accuracy and completeness because they involve door-to-door comparisons of the registers against resident citizens (Electoral Commission 2019a). Table 4.1 provides estimates from the Electoral Commission15 which found that between 8.3 and 9.4 million people in Great Britain who were eligible to be on the local government registers were not correctly registered on the December 2018 registers. They have also estimated that there were between 4.7 and 5.6 million inaccurate entries on the local government registers.

Completeness Accuracy

Parliamentary 85 % 89 %

Local government 83 % 89 %

Table 4.1: Estimates of the accuracy and Completeness of the Electoral Register in 2018. Source: Electoral Commission.

This methodology is expensive and time intensive, however, and only captures periodic snapshots of the state of the registers. Other sources of information are therefore needed to ascertain some of the longer-term trends. There is a clear picture that there has been a long-term trend towards citizens increasingly not being registered in the UK, and that this problem is getting worse (also see: Rosenblatt, Thompson, and Tiberti 2012). Figure 4.1 compares the estimates of the eligible population to the number of people on the electoral roll. This is an imperfect measure of electoral registration rates,

4 | Nine Problems with the Status Quo

‘The biggest issue is the gaps in the electoral register. It is

not just that there are people not on the register, who should be, but that under registration

is not equally or evenly distributed across society.

So obviously, there is a problem there in terms of

who is able to vote.’

Jessica Garland, Electoral Reform Society


compared to undertaking door-to-door comparisons of the names on the electoral register against the resident citizens. It under-estimates the number of missing names because it doesn’t account for inaccurate or double entries. It does provide a more longitudinal picture of trends, however.

Figure 4.1: The gap between the estimated eligible electorate and registered voters

4.4 Figure 4.2 shows the rise in the number of attainers that are thought to be missing from the electoral register. Attainers are citizens who are not yet eligible to vote, but who are entered onto the electoral register because they will become enfranchised in the near future. Attainers will therefore be approaching 18 for parliamentary elections but could be as young as 14 in Scotland and Wales for local and Scottish parliament / Senedd elections because of the lower franchise.

Figure 4.2: Missing attainers from the local government electoral registers

4.5 One possible driver for low levels of completeness is that citizens do not find the process to be convenient. The additional time involved in registering to vote is sufficient to deter many applications. Political science

research has very consistently shown that the bureaucratic effort involved in registering and voting affects whether citizens do so (James and Garnett 2020; Wolfinger and Hoffman 2001; Wolfinger and Rosenstone 1980). Citizens may not readily have a National Insurance number to hand when they want to register or may never get around to registering. Areeq Chowdhury from WebRoots Democracy told us that:

‘The problem with the current system is there’s some other barrier to people who are voting. So you think about an ideal democracy, there should be as few barriers as possible so people are going out to exercise their vote.’

4.6 Concerns were raised that the recent move towards IER would reduce levels of voter registration. Rather than a parent, spouse or household member being able to register, every citizen had to do this themselves. It was thought that younger people and students would be amongst the most adversely affected. The two-stage process to registration was also thought to make the process more bureaucratic. Academic research showed that some citizens, who were removed from the electoral register as a result of the phasing in of individual electoral registration, were turned away at the Brexit referendum, which was the first major national electoral event after the transition (James and Clark 2020). Short-term mobilisation efforts such as voter registration drives and media coverage, may have softened some of the effects, however young people and students were thought to have seen their registration rates drop (James 2020, 199-220). Taidgh Pledger from the National Union of Students told us that:

Individual electoral registration (IER) continues to present considerable challenges, particularly in terms of getting transient student populations registered to vote.

Marginalised groups

4.7 Eligible voters who are missing from the electoral register are not evenly distributed across the population. This is important because it means that some groups may


receive less political representation when electoral boundaries are drawn. It also means that they may be less likely to take part in any election. Table 4.2 uses data from the Electoral Commission’s completeness and accuracy studies to show the most significant gaps in under-registration. The table is ranked by the gap between the lowest to highest category for each socio-demographic profile. The table illustrates huge gaps by age, length of residence and housing tenure. The most under-registered groups are younger, recent movers and private renters. There are also major gaps by nationality with Commonwealth and EU citizens much less likely to be registered.

4.8 Table 4.2 also illustrates that gaps by ethnicity, disability, local authority area and SES group are strong as well. There are fewer patterns by educational attainment and urban/rural split. There was no gap by gender, although women have previously been found to be less likely to be registered.

Lowest Highest Gap

Age 16 and 17 year olds (25%)

65+ (94%) 69

Length of residence

Moved within last year (36%)

Remained in property for 16 years (92%)

56

Housing tenure

Private renters (58%)

Own outright (91%)

33

Nationality Commonwealth (62%) EU (54%)

UK and Irish (86%) 32

Ethnicity ‘Other’ (62%) Black (75%) Asian (76%)

White (84%) 22

Disability Mental disability (82%)

Physical disability/condition (92%)

10

Local authority type

London borough (76%)

Metropolitan Borough (86%)

10

Socio-economic group

DE households (80%)

AB households (86%)

6

Highest qualification

GCSE (81%) BTEC (86%) 5

Urban/rural Urban (83%) Rural (85%) 2

Table 4.2: Estimates of the completeness of the local government electoral register, ranked by gap. Source: authors, based on data from the Electoral Commission.

Voters turned away

4.9 Being on the electoral register is essential in order to be issued with a ballot paper on election day. Unfortunately, many citizens think that they are registered when they are not. A common research finding from poll worker surveys is that many citizens think that they are on the electoral register because they interact with other government services and they assume that the appropriate information is passed onto EROs.

4.10 Surveys of poll workers at electoral contests in 2015, 2018 and 2019 repeatedly found that the most common problem was citizens asking to vote who were missing from the electoral register. Roughly two thirds of polling stations are thought to turn away at least one voter at general elections (Clark and James 2017), and half at local elections (James and Clark 2019).

4.11 Meanwhile, at the 2016 Brexit referendum, the first major electoral event following the transition to IER, some citizens found that their names had been removed under the new system as they arrived at the polls (James and Clark 2020).

Funding crises4.12 There is considerable evidence that

registration teams are under considerable resource pressure. The whole of the public sector has faced squeezes on funding since the financial crisis of 2007-8 and elections have not been exempt from this. The pressure within electoral registration offices has been even greater, however, because the move to individual electoral registration brought about a much more cost intensive system for electoral officials to administer. Research shows that there was a substantial increase in staffing costs, IT costs, stationery costs and postage costs. These came from the two-stage process involving both the distribution of HEF and IVR forms (James, 2020: 214-15).

4.13 Local authorities must formally provide EROs with the resources that they need to perform their statutory duties. But in practice the provision of resources is subject to negotiation and discussion against competing


needs within the council. Alarmingly, 47 per cent of electoral officials disagreed that they had sufficient funding available to support the work required to compile the electoral register at the 2016 Brexit referendum (James and Clark, 2020). Research has also shown that local authorities cut back on voter outreach work because of financial austerity (James and Jervier, 2017).

4.14 Direct enrolment therefore poses the opportunity for local authorities to save money chasing citizens to register, who they could simply add to the electoral register with the information that they have at their fingertips, or could have at their fingertips.

Last-minute pressures

4.15 Applications to register to vote have increasingly become seasonal. Many electors apply in the immediate weeks, but usually days and hours ahead of the registration deadline. This can create an enormous pressure on electoral officials and the overall infrastructure. At the 2016 Brexit referendum the voter registration website crashed under the weight of applications. A system of automatic registration which could have a higher level of underlying completeness could help to take the pressure of electoral officials and build greater robustness into the system.

4.16 Research has shown very high underlying rates of stress for electoral officials in the UK with many having considered or having left the profession in recent years. This can have a direct effect on the quality of the election, as well as being important for members of staff themselves (James 2019).

Duplicate applications

4.17 There are an enormous volume of duplicate applications to register. This is partly because citizens are not able to check whether they are already registered using an online ‘am I registered?’ website which is often provided in other countries such as Ireland. Instead they have to contact their ERO by phone or email. Many therefore register again ‘just in case’. Only 6 percent of electoral officials reported no problems with duplicates at the Brexit referendum (James and Clark, 2020).

Use of the register by credit reference agencies

4.18 A number of respondents to our report flagged the problem with the use of the electoral register by credit reference agencies. One issue is the low fee that they pay, proportionate to the commercial gain. A full copy of the electoral register is provided to credit reference agencies, but the fee that they pay EROs is negligible compared to the value of the register for their business transactions. The fee for a data format copy of the register is set out in regulations as being limited to £20 plus £1.50 for each 1,000 entries.16 This would mean that for a parliamentary constituency of 75,000 citizens, they would pay £132.50. For an electoral register the size of the UK parliamentary register at the 2019 general election, the cost would be £71,401.17

‘A big area of concern is that when data goes off to those commercial organisations, they also get information as to who has opted out.

Peter Stanyon, Association for Electoral

Administrators

4.19 A second issue is that credit reference agencies can merge together local electoral registers to produce one single national register. There is an irony that they therefore are capable of having a better overall view of the completeness and accuracy of the register than electoral officials. This remains a potentially very useful tool for EROs, who do not currently have access to the register outside of their own jurisdiction.

16 https://www.electoralcommission.org.uk/sites/default/files/2019-11/Part%204%20Maintaining%20the%20register%20throughout%20the%20year_0.pdf p.175

17 = (£1.5*(47587254/1000))+£20. Based on the number of registered electors reported by the BBC: https://www.bbc.co.uk/news/election/2019/results


4.20 It could be argued that credit reference agencies should not be provided with the electoral register at all and that its use should be limited to electoral purposes. One local authority official told us that:

‘We keep our data safe, however we are required to provide that data to credit reference agencies who seem to be completely un-monitored in their actions.’

4.21 There is evidence, however, that many citizens register to vote to improve their credit rating rather than participate in an election. Removing the link with the credit reference agencies may therefore reduce levels of completeness and accuracy in the electoral register.

• Recommendation #1: The fee that credit reference agencies pay for access to the full register should be increased in line with inflation since 2001. The case for a higher fee to fund electoral registration modernisation should be considered.

• Recommendation #2: The sale of electoral registration data by credit reference agencies to third parties should be strictly prohibited.

The use of the edited register

4.22 The edited/open register is available for purchase by any individual, company or organisation without restriction for the purpose of the use. The edited register is commonly used by organisations who want to validate the details of their members. It is also used for marketing purposes by companies wanting to send unsolicited promotion/marketing materials. These might be considered useful for promoting commercial enterprise within the UK.

4.23 There is public value in companies being able to reach potential customers, however there are also concerns about privacy and misuse. The edited electoral register contains a rich set of data about individual citizens, data which can be particularly useful in the ‘big data’ era as it can be aggregated with other data sources and used to target individuals for political or other purposes. One electoral official told us that:

‘Laws on who can view and use the electoral register could be tightened. There are currently no criminal record checks done for those that are able to use it.’

4.24 The Electoral Commission and Association for Electoral Administrators are amongst many who have argued for the abolition of the edited/open register (White and Horne 2014). Its abolition was also called for by the Data Sharing Review in 2008 (Thomas and Walport 2008). Areeq Chowdhury from WebRoots Democracy told us that:

To be honest, I think the electoral register should just be for elections generally. I think that when people sign up on the electoral register they just want to vote. They don’t want their information being shared with some other third party.

1 sub

How 192.com use the open/edited register

192.com has compiled a citizen database made up from sources including the open electoral register; insolvency register; births, marriages & death records; land registry and business listings databases.

The database allows users to search for the personal information of citizens. It offers customers a detailed 27-page ‘background report’ on an individual, which includes data from the electoral register from 2002-2020. Using the electoral register it provides information on an individual’s name, address, likely age, length of residency, other household occupants, previous household occupants and neighbours details.

192.com also stores information about which users have searched for a citizen, including their IP address. This information is then sold.

192.com has been criticised for facilitating stalking, enabling fraud and undermining privacy.

The company has been cited as making £6.5m a year in the past.

Sources: (192.com 2020; Benett 2013; Carr 2016; Hencke 2011; White 2012)


4.25 While concerns about the use of the open/edited register have been long held, the increased simultaneous capturing of citizens’ information from their social media profile, browsing history and use of other digital applications has been raised as a concern as they have become more prevalent in society (Zuboff 2019). The combining of this data with the edited/open electoral register therefore presents increased opportunities for fraud and the undermining of privacy. 192.com is an example of a company that makes data available from the edited/open register on the internet in combination with other sources. This type of platform has often been argued to have undermined privacy (Hencke 2011) (see Box on p.26). There are also potential issues related to the political use of this data by commercial organisations – the kind of work done by Cambridge Analytica and others can be enriched and be made more powerful through the use of electoral registration data. Whilst the precise impact of this kind of work to date has yet to be fully determined, the potential for its use for electoral manipulation is significant and will grow in the future. Guarding against it is therefore important.

• Recommendation #3: The Open/Edited electoral register should be abolished.

Anonymous registration

4.26 Anonymous registration is vitally important for citizens who want to take part in the electoral process, but whose circumstances mean that they do not want their name to appear publicly on the register. This would include, but is not limited to, domestic abuse victims. Women’s Aid campaigned between 2016 and 2018 for the restrictions to anonymous registration to be relaxed. They previously required either a Court Order, proof of a Protection Order, or an attestation from either the head of the Local Authority, the head of the Police or the head of MI5. ‘This obviously wasn’t very accessible for many survivors of domestic abuse’ Lucy Hadley of Women’s Aid told us.

4.27 Changes were then brought forward in regulations in February 2018 which updated the Court Orders you can provide as evidence. It also expanded the list of

attesters that could be used to include lower level police officers, health professionals, and refuge managers.

4.28 Problems remain, however, because anonymous registration is limited to a period of only 12 months. Lucy Hadley from Women’s Aid told us that:

‘Abuse obviously doesn’t end when a relationship ends and most victims will experience continued abuse and being frightened from the perpetrator after the relationship has ended. Including through child contact processes, divorce proceedings and things that carry on after you might have left the abusive relationship.’

4.29 An alternative that has been proposed by Women’s Aid is therefore at least a five-year period, but ideally life. They also argued that the range of attesters should be further broadened to ensure it reflect the various professionals and organisations survivors may disclose to. These appear to be important recommendations given the criteria of data control set out at the start of the report and the privacy and related issues identified in Chapter 6.

• Recommendation #4: The range of attesters for anonymous registration should be expanded to reflect victims’ lived experiences of abuse.

• Recommendation #5: Anonymous registration should be expanded to at least five years.

Summary4.30 There are at least nine major challenges

with the current system. These include low levels of voter registration, strains on the administrative system and the compromised privacy of voters. The introduction of automatic re-registration stands to make a major difference to the costs involved in the system and will allow EROs to focus their resources. However, there remains a need to consider further reforms.

28 | The Overseas Experience18 http://aceproject.org/ace-en/topics/vr

19 The countries selected had the highest ranked score on the ‘Liberal Democracy Index’ for 2018 (Coppedge 2019)

5.1 By looking at how other countries run electoral registration, it is possible to have a greater sense of available options for reform. By looking at their experiences of AVR, we can learn more about what it might deliver in the UK and what risks are involved.

How electoral registration works overseas

5.2 It is possible to envisage three broad alternative types of electoral registration systems that could be found around the world, characterised by the extent to which the voter must be proactive in registering themselves, or they need take no action. There are other differences in administrative systems,18 but from the perspective of the voter, this is critical. The distinctions are therefore between:

• Laissez faire – where it is an individual responsibility to register and the government only provides the minimal infrastructure to help citizens.

• Assisted – where there are regular prompts to citizens, encouraging them to register to vote when they interact with other government services.

• Automatic - where a citizen needs to take no action in order to be registered to vote – it is done by governmental bodies on their behalf.

5.3 Table 5.1 opposite summarises the position of 40 countries considered to be liberal democracies around the world.19 There is no single consolidated list on which system is used and so a bespoke dataset was established, in collaboration with Professor Rodney Smith, University of Sydney, using election observation reports, official websites and other sources.

Laissez faire Assisted Automatic

Cyprus Australia (varies by state) Argentina

Ireland Canada (varies by region) Austria

Jamaica USA (varies by state) Belgium

Mauritius Chile

New Zealand Costa Rica

United Kingdom Czech Republic

Denmark

Estonia

Finland

France

Germany

Greece

Iceland

Italy

Japan

Latvia

Lithuania

Luxembourg

Netherlands

Norway

Panama

Portugal

Slovakia

Slovenia

South Korea

Spain

Sweden

Switzerland

Taiwan

Trinidad and Tobago

Uruguay

Table 5.1: The use of AVR around the world

5 | The Overseas Experience

The Overseas Experience | 2920 https://www.osce.org/odihr/elections/latvia/387665?download=true 21 https://www.cvk.lv/uploads/files/rus_eng/Electoral%20Register%20Law.pdf

5.4 Most countries do have some kind of automatic system. This is much easier in many polities than it is in the UK because it is drawn from a central population registry. However, there is usually an option for citizens to check the register to ensure that they are on the list and have not been inadvertently excluded. In the Czech Republic registers are not public but citizens can check their inclusion and request corrections up to two days before polling day. In Denmark citizens are sent voter cards and this enables corrections to be made up to and including the day of the election.

5.5 Some countries only have automatic enrolment at particular ‘life moments’. All citizens are registered at the age of 18 in France from national records, but after that are required to notify election officials of a change of address. This is also the case in Trinidad and Tobago where citizens are registered when they apply for a national form of identification, which they can do upon reaching the age of 15. They are then transferred to the register at the age of 18. They need to inform the electoral authorities if they have subsequently moved.

5.6 There is also some variation in the data sources used to compiled the register automatically. Some electoral authorities use national civil registries such as Lithuania and Estonia. Others rely on local sources. Switzerland used canton-level civil registers, for example.

5.7 There are cases where there is no system of electoral registration. In Latvia citizens arrive at a polling station and must present their passport in order to vote. A stamp is then affixed in the passport to indicate that they have voted, to prevent multiple voting. Eligible citizens without a passport must collect a voting card three weeks prior to election day.20 Information is sent to local areas based on population registers to help organise the polls, however.21 Latvia is therefore classified as automatic since no action is needed by the voter, despite the absence of a system.

5.7. ‘Assisted’ systems are present in countries such as parts of the USA, Australia and Canada. These require an individual to apply to register to vote, but there is some integration with other public services.

5.8 Laissez faire systems such as Cyprus and New Zealand which require an application to be submitted, are very few in number. The UK system is best described as ‘laissez faire’. Voter registration is compulsory and there are sanctions for non-registration. However, these sanctions are so rarely applied that the system can be described as laissez faire. These countries all have direct connections to British imperial rule, which shows the likely source of the registration system.

5.9 An important trend has been for countries that have historically had laissez-faire systems to adopt forms of assisted or automatic voter registration because they have found long-term declining rates of registration. In the USA, sixteen states and the District of Columbia have already approved some form of automatic voter registration by 2019 (Brennan Center 2019). Canada also moved to a system of automatic voter registration (Black 2003). In the next section we profile the case of Victoria in Australia, to show how automatic voter registration has been successful for boosting enrolment amongst particular groups where there is no population register.

What effect does AVR have? Lessons from Political Science in the USA

5.10 Higher rates of voter registration do not guarantee higher rates of voter turnout. One concern about introducing AVR is that it might not be worthwhile because it would register citizens, who may then not go on to vote.


5.11 There is a long-standing stream of political science research which shows voter registration procedures can affect turnout. Much of this is based on analysis of divergent procedures in the US, where each state historically has had relatively autonomy. In a famous study, Rosenstone and Wolfinger (1978; 1980) compared registration practices to levels of participation in the US states to calculate the changed probabilities of an individual voting. They suggested that, depending on the existing probability that an individual would vote, imposing a 30-day registration deadline could reduce the probability by between 3 to 9 percentage points.

5.12 The US 1993 ‘Motor Voter Act’ forced US states to provide registration facilities at public agencies, introduce universal mail registration, allow citizens to register at the same time as applying for/renewing a driving licence and prohibited officials from purging electoral registers of non-voters. A number of early studies showed that this did increase voter participation immediately following its introduction, with one finding that it increased registration by 2.3 percentage points and turnout by 2.1 percentage points (Franklin and Grier 1997, 111).

5.13 The move to full AVR has been a more recent development in the US. Data has been collected on registrations and turnout statistics from all eight jurisdictions that implemented AVR in time for the 2018 general

election (Morris and Dunphy 2019; Rakich 2019). Through the AVR systems, 2.2 million new voters were registered and the details of six million were changed (Rakich 2019). The increase in registrations varied by as much as 9.4% in Washington DC to 93.7% in Georgia, but was always positive (Morris and Dunphy 2019, 2). Analysis has found that people who were registered through AVR do vote, but at a lower rate than those who register themselves. The turnout rate for AVR was 42% in Colorado, for example. Further research will be needed in future years to assess the effects on turnout, but the evidence so far suggests that it has had a positive effect – albeit not ‘game-changing’ in nature.

5.14 AVR requires careful and successful implementation, however, and there have been cases of problems. In the first year of implementation more than 100,000 registration errors were reported in California including 1,500 ineligible electors being added to the roll (Ernst and Young 2019; Vasilogambros 2019).

Case Study: Direct Enrolment in Victoria, Australia

5.15 AVR has also been partially introduced in Australia. The remainder of this chapter provides a more detailed case study of the effects of the reforms there, focussing in on the Victorian Direct Enrolment programme. Australia has a highly federalised system in

The Overseas Experience | 31

which the Australian Electoral Commission (AEC) is responsible for organising and supervising federal elections, but state and local government elections are overseen by a separate Electoral Commission in each state and territory. There is therefore a federal electoral register held by the national Australian Electoral Commission, and a register for local and state elections.

5.16 Electoral authorities traditionally relied on ‘habitation reviews’ for voter registration. These were similar to the UK’s annual canvasses, designed to maintain the electoral register. As early as the 1990s, however, electoral commissions across Australia began to use data from government and governmental agencies to write to people likely to need to enrol or change their enrolment status. This included the sending of a birthday card to citizens aged 17 inviting them to register (Victorian Electoral Commission 2017, 1). It also included using data from Centrelink, Australia Post, driving licence data and a variety of state/territory level data sources to write to citizens.

5.17 The datasets were initially helpful in identifying people to remove, but unless citizens responded to the new invitation then more names would drop off the register than would be added on. A decline in the response rates to these invitations therefore led to calls for automatic enrolment (Victorian Electoral Commission 2017, 1). Direct enrolment first came to New South Wales after the NSW Parliament passed legislation to enable SmartRoll in 2009. Victoria followed in 2010 and the Commonwealth Parliament in 2012 (Reader 2015).

5.18 The NSW SmartRoll system was praised by a cross party parliamentary group as a ‘considerable achievement’ (JSCEM 2012, 39), a few years after implementation. When citizens updated their address with

another NSW agency (driving licence data, schools data, and home ownership data), this information was passed onto the NSW electoral officials who could then automatically update the electoral roll. Electors were then notified by SMS, email or letter. Of the 42,172 SmartRolled voters ahead of the 2011 NSW elections, 72% exercised their right to vote (JSCEM, 2012: 58). The scheme was ended in 2019, however, as the state moved to use alternative data sources provided by the national AEC.

5.19 At the federal level direct enrolment involved data being transferred from public agencies such as the State and Territory’s Driver’s Licence Authorities and Australian Tax Office. Citizens would be added to the register or have their details changed if the AEC decided that it had sufficient information to be confident of the change. Citizens would then be written to and would have 28 days to respond if the details were incorrect (AEC 2019).

The Victorian Direct Enrolment Scheme

5.20 The Victorian Direct Enrolment scheme, focussed on in more detail here, involved enrolling new electors, and not updating current registrants. Citizens would be automatically added to the register 21 days after an elector turned 18. The programme began by auto-enrolling and writing to nearly 2,000 students who were registered with the educational authorities but who were missing from the register. This was expanded to using data from the Department of Immigration and Border Protection, as well as data on death, births and marriage records. A legislative amendment was passed in 2013 which permitted the Victorian Electoral Commission to automatically enrol citizens who were on the AEC’s federal roll.

5.21 The number of new direct enrolments reached 191,849 per year in 2015/16 (Victorian Electoral Commission 2017, 2). These new enrolments were more likely to be male

Direct enrolment in Victoria was

very effective at encouraging voting for first-time voters


(56.8%). The majority (45.8%) were in the 18-19 age category, with many others in the 20-24 (22.9%) and the 25-29 (13.2%) category (Victorian Electoral Commission 2017, 3-4).

5.22 There was concern that direct enrolment might simply add citizens onto the register who might not vote. There was some evidence of this. A notable finding was that 61.7% of direct enrollers cast their vote in the 2014 state elections, compared to 93.0% of all enrolled citizens. It was very noticeable, however, that 79.4% of 18-19 year old direct enrollers did cast their ballot. It was the older age groups where direct enrollers were less likely to participate (Victorian Electoral Commission 2017, 11-2). Importantly, direct enrolment was therefore a very effective at encouraging voting for first-time voters.

5.23 Feedback from directly enrolled electors showed that relatively few (7%) had deliberately not enrolled themselves. The most common reason for not registering themselves was that they had assumed that the VEC would register them (42%), that they were going to get round to enrolling (30%), that they had forgot (13%), didn’t know how (14%) or didn’t know that they needed to (13%). The vast majority were in favour of direct enrolment (82%) with a small proportion (4%) disagreeing with it (Victorian Electoral Commission 2017, 18). Concerns about privacy were raised from time to time. As one official from the VEC told us:

‘We often get concerns about privacy: how do you get my information, where did you get it from? Why are you using it to update my enrolment? Those questions come fairly regularly.’

5.24 Mechanisms for improving the transfer of the data improved considerably over time to reduce cyber security threats. The VEC was originally using CD-ROMs and USB sticks for data transfers. A private cloud service was subsequently set up.

5.25 No public estimates have been published about the net cost effect involved in introducing AVR. One official told us, however, that direct enrolment provided 10-15 percent of the enrolment in Victoria. Without direct enrolment, there would be the need to chase eligible citizens and there would be 10-15% more time spent processing.

Summary5.26 AVR is the norm, not the exception in

countries around the world. Many countries that have historically not had AVR because of the absence of a population register are now increasingly turning to introduce either direct enrolment for specific groups or assisted voter enrolment through other public agencies. These innovations, where they have been well-designed, have proven to be able to deliver cost savings and boost voter registration for specific groups. Successful implementation is reliant on identifying reliable new data sources, training, resources and project management, however.

1 sub

AVR is the norm, not the exception in countries around the world. Many

countries that have historically not had AVR because of the absence of a population register are now

increasingly turning to introduce either direct enrolment for specific groups or assisted voter enrolment

through other public agencies.

34 | Philosophical Choices and Issues

The report has so far identified some areas of critical strain in the UK’s electoral registration system, and that some form of AVR is the norm in liberal democracies around the world. There are a number of philosophical questions to be considered when contemplating the case for AVR, however, which this chapter aims to introduce.

The case for voluntary voter registration

6.1 The most obvious question is whether voter registration should be opt-in, or opt-out. One common argument that has been raised against AVR is that it should be an individual’s responsibility to register themselves and/or that an individual should have a right not to be registered. These arguments have commonly been raised where AVR is being proposed. For example, the Liberal Party opposed AVR in New South Wales, Australia, stating that ‘we do not believe that it is the right of the State to put people on the register’ (JSCEM 2012, 38). In a debate on AVR in the UK Parliament in 2016, meanwhile, the Minister for the Cabinet Office, John Penrose, raised the concern about whether automatically adding citizens to the electoral register contravened civil liberties:

‘There is a difference between an opt-in system, where we say, “We know you’re living there, but do you want to register?” and an opt-out system, which is one possibility, or a “we’re not even going to ask you” system, which is a bit more dangerous. Whether that is really acceptable in a free society is a bit more questionable; it is tricky in some respects from a civil liberties point of view.’ (Hansard 2016, Volume 612, Column 146WH)

6.2 The Coalition government originally held voter registration is an act ‘of personal choice and as such there should be no compulsion placed on an individual to make an application to register to vote’ (Deputy Prime Minister 2011, 20-1). It therefore proposed to abolish the fine for the non-completion of an electoral registration form.

The case for mandatory voter registration

6.3 One of the arguments for mandatory voter registration is that voting is a positive act, from which all individuals gain, and it should therefore be promoted by the state. If a citizen casts a ballot then their voice is more likely to be heard by government, parliamentarians and other elected officials (Dahl 1971). Voting can also build civic education as citizens consider who to vote for at an election (Sheppard 2015). One person voting can also lead to others voting (Nickerson 2008). The state therefore has a responsibility for taking action to encourage everyone to vote (James and Garnett 2020).

6.4 Another argument for mandatory voter registration is that at the collective level, citizens mutually gain from higher participation. Higher turnout can increase confidence and legitimacy in the result of an election.

6.5 A further argument in favour of mandatory

voter registration is that, without it, electoral registers and election outcomes will not be representative of the public. The ACE Electoral Knowledge Network explains that:


22 http://aceproject.org/ace-en/topics/vr/vra/vra10 Philosophical Choices and Issues | 35

Inevitably some voters will choose not to register, thereby effectively depriving themselves of the right to vote. If there is a random distribution to the likelihood of registering to vote that is, if all types of citizens register in equal proportions, whether men or women, young or old, urban or rural, rich or poor, highly or poorly educated, and so on the conclusion might be that voluntary registration has no impact on the outcome of an election or on the selection of representatives and governments.22

6.6 Electoral registers are also currently used for purposes beyond conducting an election, such as the drawing of electoral boundaries, the detection of crime and jury service. If voter registration was optional, then some groups would be less likely to be represented in Parliament and the judicial process would be less representative.

6.7 It is already current law that every citizen is registered. Voter registration is therefore not currently an ‘opt-in’ process. This report progresses on the assumption that this continues to be the case. AVR is therefore a

policy to help administratively best realise the current goal of full, compulsory registration. However, voters should be aware of how information about them is being shared. There is therefore a case for an ‘opt-in’ when registering via other government services.

• Recommendation #6: Voter registration should remain compulsory across the UK

Automatic or assisted voter registration?

6.8 Using a variety of non-electoral data sources to register citizens raises a separate question about whether citizens need to grant permission for this information to be transferred. Should a citizen be asked, for example, when they pay their Council Tax whether they should like to register to vote at that point? As the previous chapter outlined, this would be assisted registration. Or should they not be granted the opportunity to be asked and automatically added to the roll? This would be automatic registration.

6.9 One argument for asking them – an ‘opt-in’ system – is that they should retain greater control of their data. If direct registration was introduced without the consent of the voter, then their details might be passed onto credit reference agencies or even third parties via the open/edited register, without their full understanding.

6.10 An argument against asking for consent, however, is that they may decline. Administratively cost-effective ways of boosting the accuracy and completeness of the electoral register would therefore not be achievable. For example, people might avoid supplying if they became concerned that the data might be used in other ways, not just amending the electoral register but making the data available to others via the electoral register. This is another reason to restrict the use of the register to what is strictly necessary – avoiding commercial or political use of the register – as it would reduce this possible chilling effect.

• Recommendation #7: If AVR is introduced, then directly enrolled citizens must not have their details added to the open/edited register (should that register not be abolished).

Does compulsory voter registration contravene civil liberties?

The question of whether a person has a right not to be registered to vote could be seen as a civil liberties issue but in most ways this is a stretch as the civil liberties issues surround the way the electoral roll could be misused rather than used for its primary function. If the register can only be used in relation to actual voting the impact upon civil liberties is effectively nil.

The problems arise where the register is used for other purposes – whether intentionally for political or other analyses or for marketing purposes to raise money to fund the register – or unintentionally through function creep, data breaches or worse, or through the potential use for issues such as immigration enforcement. Avoiding these kinds of uses would be a civil liberties issue – so the key is legal uses and by ensuring security of the register is taken as seriously as possible.


Privacy and data protection issues

6.11 AVR raises some important privacy and data protection questions. At the outset, however, it is important to note that privacy issues and data protection issues are not the same. There are aspects of privacy that are not related to data protection – the ‘chilling effect’ that means that knowing you will be on a list in the hands of the government and others could have an influence on whether you perform some task or avoid some key activity, for example. There are data protection issues that are more based on the way the law works than on privacy itself. When considering privacy, therefore, it is important not to reduce the process to a ‘compliance tickbox’ related to the precise terms of the GDPR, but to take a broader look at the implications of what is being suggested.

6.12 Privacy issues can occur at each stage of the process. At the data gathering stage, there are privacy issues when considering whether a particular data source might be used. Would the knowledge that the data source could be used for an electoral database change anything about that data source itself? Would people be less willing to contribute to it? Is the use of it for an electoral register consistent with the way in which the source itself was compiled? Would its use for an electoral register be expected or understandable for those whose data is being considered? Should the people whose data is being gathered be informed, or be given the chance to opt-out of their data being gathered from this source? These are practical, ethical and potentially legal issues – in different circumstances they could bring in the lawfulness of the use of the data from a GDPR perspective into question.

6.13 There are privacy issues in relation to the holding of data. These include security issues – these kinds of issues are relevant to any government database, but if the data is held

in a particular structure and concentrated way they can become more significant. A larger database is a larger target. One that is guaranteed to be especially accurate and reliable, as an electoral register would have to be, is potentially more valuable to hackers and others. One that includes specific location data, as well as personal details such as age and nationality, is also correspondingly of more potential value to those hackers.

6.14 The point is not just that the data is valuable on its own – for marketing, for potential targeting of customers or victims of scams, for demographic profiling and potential political manipulation – but that it can be

combined with other data to make that data more valuable and more risky. For example, if social media data has been scraped and analysed to find online activity of people with particular views, but those online identities can’t be matched to real people in the real world, getting hold of the electoral register could allow the data to be matched and those online identities located in the real world. Conversely, if the electoral data allows you to find people in a particular

‘swing’ constituency, it could be matched up with that social media political activity data to find persuadable people in that swing constituency to try to change their vote, make them more or less likely to vote and so forth. This is just one example in the political sphere – the potential uses in this and other spheres are immense.

6.15 All these factors mean that the security issues on an electoral database are particularly significant, especially if data is to be combined into a single, central register, as discussed below. Once data has been hacked or lost, the privacy issues grow - because the control over the data has been lost, and how it can be used, where else it could be sold or made available, is no longer in the hands of those who should have been responsible for it.

The data is valuable on its own but can

be combined with other data to make that data

more valuable and more risky.

Philosophical Choices and Issues | 37

Possible data misuse issues

6.16 Perhaps the biggest of the privacy issues, however, relate to the possible misuse of the data. Misuse can come in a number of forms:

• Accidental misuse – where the data is used for something that the user does not realise is not acceptable or appropriate.

• Intentional misuse – where the data is used for something that the user realises might not be appropriate but does it regardless.

• Function creep (or mission creep) - where a new use is found for data, one that was not originally envisaged but emerges later, perhaps as ideas change, or attitudes change, or in response to a particular event.

6.17 The first issue, accidental misuse, is best guarded against by well-written policies and well-trained staff, as well as clear limitations on the planned use. The more limited the allowable uses, the easier it is to see that another use would not be appropriate. This also helps guard against intentional misuse, as it reduces the potential grey areas so discourages moving into them. Again, well-written policies are key here, together with appropriate penalties and procedures to deal with breaches.

6.18 Function creep/mission creep is perhaps the hardest to deal with. This happens with regularity in most fields where data is concerned. For example, the ANPR camera system designed for the London Congestion Charge was originally intended just for that charge, but soon began to be used for the detection and prevention of crime. Function creep generally happens for what appear to be good reasons at the time – sensible ideas, efficient uses of resources, ways to solve difficult or immediate problems – but that does not mean that it is necessarily a good idea in other ways. It means that assessments of privacy risk, for example, may need to be reperformed. It means that consent that was originally given for the use of data may no longer be valid. It also potentially leads to further function/mission creep. One new idea leads to another. A database like the electoral register could have many such possible uses – and many associated risks.

6.19 There are two keys to guarding against function creep. The first, as for the other

two key forms of data misuse, is to limit the allowed uses from the start as tightly as possible. If no commercial use of the electoral register is allowed, for example, then the temptation to look for further uses disappears. The second is to be aware of the possibility of function creep, so to refuse even what seem to be very sensible and appropriate possible additional uses of the database. This is much harder to do as the idea of getting the maximum value from a database is a strong one, and the idea that sharing data is essentially a good thing is compelling. It is, however, in privacy terms a very risky one.

Is there a need for a data trust law?

6.20 Rather than a specific law being needed to deal with potential misuse of electoral databases, the provisions of existing data protection law should be applied strictly and carefully. If the precise uses to which databases might be used are set out clearly and restrictively then data protection law should mean that inappropriate use is prevented.

6.21 There are complications to this, however, that need to be monitored closely. One is the impact of Brexit on the way that the GDPR applies in the UK, which is not yet clear. It has been suggested that after Brexit, data protection laws might be weakened – this might have an impact on electoral databases. If this happens, specific restrictions should be written into any revised data protection


law. The second is how the exemptions from the GDPR built into the Data Protection Act 2018 are used in practice. There is a specific exemption for immigration (Schedule 2 (4)) in particular that needs careful monitoring. An argument could be made that the register could be used for immigration enforcement – if you are not on the register you could be assumed to be worthy of investigation, for example. This would be function creep/mission creep and should be avoided not just for its immediate impact but as a precursor to further uses for ‘enforcement’.

• Recommendation #8: Legal restrictions on the use of the electoral register should be considered should GDPR laws change following Brexit.

Data security and cyber threats

6.22 There are inevitably cyber threats involved in voter registration systems. As the Center for Internet Security notes:

‘In general, voter registration systems exhibit the risk characteristics of a general-purpose computing system and, more specifically, any network connected database application. To properly mitigate risks, each voter registration system… and links to the voter registration system, needs a comprehensive assessment of its technical characteristics and the application of appropriate security controls’ (Center for Internet Security 2018, 16).

6.23 Cyber security weaknesses are important because they can lead to the leaking of personal information which can undermine privacy, as this chapter has already considered. It can also lead to problems such as tampered registers which could enable electoral fraud, or the deliberate sabotage of electoral registers which could prevent elections from being held reliably and on time.

6.24 The strength of the election infrastructure in a given country to provide election security reflects the range of actors available to redress and respond to cyber security breaches (Brown et al. 2020, forthcoming). In this sense, the UK is well equipped with a range of bodies including the National Cyber Security Agency, in addition to the Electoral Commission, Cabinet Office, local authorities,

private sector suppliers and a strong university network.

6.25 The current separation of the UK’s electorate into multiple local registers also provides a degree of increased security since the data is dispersed and there are fewer connections than there otherwise would be with a single centralised dataset. There are no known high-profile breaches of the UK electoral registration system that the authors are aware of, compared to the US (United States Senate Intelligence Committee 2019a, 2019b). At the 2016 Brexit referendum, however, the voter registration website crashed, and a targeted distributed denial of service was not ruled out as a potential reason why the site was temporarily unavailable (The Public Administration and Constitutional Affairs Committee 2017).

6.26 The Electoral Commission has already undertaken feasibility studies of the automatic transfer of data using the IER Digital Gateway and found it to be a reliable system (Electoral Commission 2019d). Interviews and written submissions to this report provided no evidence that there would be a greater threat of cyber security as a result of transferring more data through the IER Gateway – since secure information already passes through this route. As one administrator claimed: ‘Local government already has a lot of information on citizens for tax purposes, this would be no different.’ The move to AVR would increase the number of connected datasets and the security of each would have to be carefully assessed. The strong network capacity of the election sector, however, should mean that delivering AVR securely would be possible.

Centralised register

6.27 The UK government had proposed the implementation of a single electoral register with the Co-ordinated On-line Register of Electors (CORE) project, which was legislated for in 2006. This was repealed in 2013, however, because of concerns about cost (see paragraph 2.4). Wales plans to introduce a Welsh Database of Electoral Registration Information of local government electors.

6.28 The advantage of a centralised register is

Philosophical Choices and Issues | 39

that it would enable EROs to be more readily able to check for the duplicate electors in the register. The register would involve each citizen having a unique electoral identifier. Should they receive a wider set of data to update the register or direct enrolments, as sketched out in this report, then they would be able to check when/where they were registered elsewhere. Lastly, EROs would be more efficiently and securely able to transfer data to other EROs for the purposes of electoral events rather than exporting data through locally held files.

6.29 The main disadvantage is that the reliance on one single register for all electoral events would create a data ‘honeypot’ which hackers may want to target. Pascal Crowe from the Open Rights Group said that:

‘You’d have to be very careful about how you stored that data because if you are storing it all in one place then essentially the risk of it getting hacked is higher because, obviously, a big database is worth more to hackers than lots of small databases.’

There would also be costs involved in setting up the register, which was one of the main reasons why the CORE project was cancelled.

Organisational reform

6.30 Is there still a case for having local EROs responsible for the electoral registers in

their area? Or is there a case for transferring responsibility to the Electoral Commission? Research shows considerable value in having local officials in control of the registers. They have local knowledge of buildings, properties and local cultural/socio-economic factors that puts them well positioned to undertake the canvass and maintain the register (James 2017). Responses to interviews were also broadly supportive of local officials retaining control of their registers.

• Recommendation #9: Local EROs should continue to be the main custodians of the electoral register in the geographical area, with support from the Electoral Commission.

Summary6.29 The decision as to whether to make

voter registration compulsory or not is a philosophical one. A philosophical objection to AVR is that individuals have a right not to be registered. Democratic theorists have argued, however, that measures are needed to create a level playing field at an election. There are also privacy and other issues raised by AVR, which reinforces the need for the register not to be publicly available for sale.

1 sub

There are multiple options for building automatic and/or assisted voter registration into our democracy.

Options for Implementing AVR | 41

7.1 The report has so far identified problems with the existing voter registration system, the normative case for reform and experience from overseas. But is AVR actually achievable in the UK? What options are there? What might the consequences be, beyond the completeness and accuracy of the register? Five overlapping options are posed here.

Approach #1: A civil population register

7.2 AVR is used in many countries because they have population registers. Electoral registers are simply then constructed from a transfer of data from that registry ahead of the election. Citizens and political parties are then usually given an opportunity to check the register ahead of election day. The construction of a new population register would certainly simplify voter registration in the UK, could make it less resource intensive, and easier for the citizen to understand. Luke Butcher from the South Korean National Electoral Commission told us that:

‘In my view automatic voter registration is going to require the UK to have a computerized civic register with ID cards. Until this happens, I am not sure there will be anyway to do automatic registration effectively.’

7.3 The Labour government passed the Identity Cards Act 2006, with Charles Clarke as Home Secretary, that would establish a National Identity Register (NIR) and a national identity card. ID cards were seen as a way to reduce crime and track people coming in and out of the country (Labour Party 2005).

7.4 This was subsequently repealed by the Conservative-Liberal Coalition government in the Identity Documents Act 2010. Theresa May, as Home Secretary introduced the Bill arguing that:

‘The national identity card scheme represents the worst of government. It is intrusive and bullying, ineffective and expensive. It is an assault on individual liberty which does not promise a greater good’ (May 2010).

7.5 Conservative MP Jacob Rees-Mogg was quoted as saying that:

‘ID cards change the relationship between the individual and the state. We have a right to go about our business and not be stopped and asked where we are going. If you have ID cards a policeman can ask you at any time who you are and what you are doing’ (The Week 2018).

7.6 Civil rights campaigners claimed that they involved an encroachment of civil liberties, helping usher in a ‘your papers please’ culture where people have to ‘prove’ their entitlement to what should be available to all without question. It is easy, by this argument, for people to feel excluded through the enforcement mechanisms that require ‘proof’. If ID cards are not to be a restriction on freedom, the functions to which they may be put should be very tightly limited, not just at inception but as time passes. The civil liberties argument is that though this is theoretically possible, in practice ‘mission creep’ is inevitable. At the very least, it needs to be guarded against very carefully – but in practice this argument may well be strong enough to mean that this kind of a system should be avoided.

7.7 There have been recent calls for a civil population register and ID cards to be reintroduced in the wake of the Windrush scandal and Brexit from the thinktank Policy Exchange (Goodhart and Norris 2018) and, former Home Secretaries Charles Clarke, Alan Johnson and Ken Clarke (Ferguson 2018).

7 | Options for Implementing AVR

42 | Options for Implementing AVR

7.8 There are no manifesto commitments to introduce a civil population register (although the Conservative government is committed to requiring voter ID at polling stations), so such a major reform would currently have little public mandate. It is likely that there would be considerable costs involved in setting up a new system. It is therefore unlikely that this policy would be pursuable at this time. Further, the political risks of mission creep might well be seen as growing at time of political turbulence, rising nationalism and related issues.

Approach #2: Use of an expanded DWP CIS database

7.9 Although there is currently no formal population register, the existing voter registration system is already contingent on checking applications against information in the DWP Customer Information System to verify electoral registration applications. An alternative would therefore be for the DWP CIS system to ‘double up’ as the electoral register, or a central elector database could be created, connected to the DWP CIS.

7.10 The advantage of this is that there is already a unique identifier for each individual in the system (a NINo). The centralised nature of the DWP CIS would allow for duplicates to readily be checked. This would bring some significant broader organisational changes:

• Actively used fields for elections in the DWP CIS would have to be expanded to include nationality and eligibility for different types of elections.

• It may require greater capacity within the Cabinet Office, the Electoral Commission or a central government department to manage the system.

• An online portal through which citizens could check their registration status might be easier to implement.

• The system could still retain EROs as the custodians of their register for their geographical jurisdictions.

This would clearly, however, be a major reform that would take long-term planning beyond the current Parliament.

• Recommendation #10: The establishment of a central electoral register based on the DWP CIS is explored in the long-term.

Approach #3: Data-mining

7.11 A third option for introducing AVR would be to maintain the existing localised system of voter registration, but allow for additional data sources to be made available for EROs, which they could then use to automatically register citizens. EROs could be given the opportunity to match various datasets and have the capacity to add names to the electoral register, if they are missing. This was proposed in the Local Government and Elections (Wales) Bill 2019.

7.12 EROs already have access to a range of local datasources, however. The extent to which they use these varies and is contingent on resources and personnel (Clark 2014, 2015). The use of performance standards has previously been used to incentivise EROs to undertake data matching. Research has found that publishing information about whether EROs undertake activities, including using local data sources, encourages compliances because of fears of reputational damage to individual members of staff or local organisations (James 2013). Performance standards could again be used to incentivise EROs to data mine and register citizens automatically.

7.13 It is questionable whether data-mining is effective, however. Data-mining pilots have already been undertaken by the Cabinet Office in collaboration with the Electoral Commission and EROs as part of the preparatory work to introduce IER. ‘Data mining’ in this context meant ‘exploring the extent to which access to information held on national public databases can assist Electoral Registration Officers (EROs)’ (Electoral Commission 2013, 1). The conclusion of those pilots was that:

‘The amount of time and resources they spent on reducing the data provided to a list of likely new electors with usable address information was unsustainable and could not be incorporated into their ‘business as usual’ processes’ (Electoral Commission 2013, 3).

7.14 A Cabinet Office evaluation also flagged concerns about the data quality in other datasets. Duplicate entries, inaccurate addresses, ineligible electors and other issues made data matching difficult to proceed with (HM Government 2013).

Options for Implementing AVR | 43

7.15 This system could be relatively quickly implemented. A system that was reliant on EROs having sufficient time and resource to check other data sources, given that they already lack the time and resources to fully utilise local data sources, may not be the best way forward, however.

Approach #4: Direct Enrolments for Specific Under-Registered Groups

7.16 A fourth approach for introducing AVR, which would also maintain the existing localised system of voter registration, would be for direct enrolment to be organised for specific groups when they undertake new transactions with government services. For example, if you were to apply for a driving licence or state benefit, then EROs could be alerted to the new transaction at the point of time at which it occurred. The advantage of focussing on new transactions is that each transaction is likely to represent a more up-to-date set of information than is held by the ERO.

7.17 Information from new transactions could be passed along the IER Gateway to EROs. It would then be received by electoral officials in their Electoral Management System, in the same way that voter registration applications are received. The applications could then be processed using the usual procedure (see Chapter 3 above). The Electoral Commission (2019d, 1) has already established that:

‘[T]echnology already employed by the Individual Electoral Registration (IER) Digital Service could form the building blocks for further reform. Further development of the IER Digital Service could see it acting as a conduit, receiving transactional data from new and reliable national data sources.’

7.18 This system would involve less radical change and could be targeted at particular groups where registration rates are low. It would follow on from the success of international experiences noted in Chapter 5.

• Recommendation #11: Direct enrolment for specific under-registered groups when they undertake new transactions with other government services

Approach #5: Assisted Voter Registration

7.18 A final approach would be that citizens were simply prompted to register to vote when they accessed other government services. Registration is only automatic when citizens opt-in for their details to also be sent to EROs. This is described as ‘assisted voter registration’ in this report. The advantage of this approach is that citizens still retain control over the process and their data will not be passed on without them being consciously aware that this has happened. A disadvantage is that citizens may still opt not to register via this route. If there are too many opportunities to register to vote in this way, then EROs might also be overwhelmed with applicants.

• Recommendation #12: Citizens are prompted to register to vote when they access certain government services.

SummaryThere are a variety of ways in which the principle of AVR could be partially or entirely introduced in the UK. Complete introduction would seem to require the move to a population register or, at least, the construction of a new centralised electoral register based on the DWP CIS. Partial implementation could be realised by automatically enrolling citizens when they complete specific transactions and this might be a more practical option. Assisted voter registration could also be realised by prompting citizens to register to vote when they access government services with ‘opt-in’ boxes.

44 | Finding the ‘Life Moments’: New National Data Sources for Direct Enrolment

AVR could improve the completeness and accuracy of the electoral register through direct enrolment for specific groups and providing assisted voter registration when citizens access other government services. This chapter identifies the new national data-sources that could be used, based on original figures compiled from Freedom of Information requests.

Criteria for data sources

8.1 There is an enormous volume of transactions that take place each year between citizens and other government bodies. The UK government has published a list of transactions for services that it provides – there were 776 in total – and the volume of such transactions per annum.23 Not all of these would be suitable for compiling the electoral register, since in many cases the applicant is a business, for example. However, there would be huge potential to use many of these data transactions to improve the completeness and accuracy of the electoral register.

8.2 There are some criteria that should be considered when identifying possible uses.

• Selectivity. Too many data streams could

make it difficult to manage for EROs. There should therefore be some selectivity – and certainly to begin with.

• Reliability. Information should be accurate and reliable – for example, the name and addresses provided should be the most accurate record.

• Completeness. The data should be sufficiently complete for EROs to be able to use to amend the electoral register or prompt the citizen to register to vote.

• Traffic volume. Data sources should bring a

significant volume of transactions for the process to be worthwhile.

• Addressing gaps. A further consideration is that the data sources should identify those groups who are least likely to be on the electoral register with the aim of addressing those gaps, and not extend any existing disparities between groups who are missing from the register.

• Preventing new inequalities. The selection of data sources should not introduce any other new inequalities in the completeness of the register between groups.

• Appropriateness. At some life moments it would not be appropriate to ask – such as booking prison visits (of which there were nearly one million transactions) because it could create a negative association for the citizen.

• Privacy and consent. Citizens should have a reasonable expectation that the relevant data-stream might be used for this purpose, and in some circumstances give consent to that use. Assessing that reasonable expectation and ensuring that any consent gained is informed will require an understanding of all the purposes to which the register will be put, which will be easier if those purposes are very strictly limited.

8.3 With the criteria of selectivity in mind, it makes sense to focus on new transactions that citizens have with governmental bodies rather than the repeated transactions that they might have since they would be more likely to enter fresh data, rather than have data roll forward. A selection of the government transactions is published in Appendix C which could meet these

8 | Finding the ‘Life Moments’:New National Data Sources for Direct Enrolment

23 https://www.gov.uk/performance/services?sortby=department_name&sortorder=ascending. In most cases the government data

that it has published was from 2016/7, so new data was requested for this report and is published below.

24 For other suggested sources, see: https://services.parliament.uk/Bills/2016-17/automaticelectoralregistrationschoolstudents.html

Finding the ‘Life Moments’: New National Data Sources for Direct Enrolment | 45

criteria.24 The next section considers the advantages and disadvantages of using each of these data sources, informed by updated information about the volume of transactions based on Freedom of Information requests.

Issuing of National Insurance Numbers

8.4 When a Child Benefit claim is made HMRC allocate a Child Reference Number to the child. When the child approaches 16 years old, the number is converted into a National Insurance number. This is known as a ‘juvenile registration.’ HM Customs and Revenue then writes to citizens, three months before their 16th birthday to inform them of their National Insurance number. Table 8.1 shows the number of such Juvenile Registrations for National Insurance numbers that there were for each tax year from 2015/6 to December 2019.25

2015/16 2016/17 2017/18 2018/19 April 2019 to December 2019

703,446 691,748 681,748 699,116 541,235

Table 8.1: Juvenile Registrations for National insurance numbers 2015/16 to December 2019

8.5 Given that 16-year olds are able to be attainers or vote in many elections across the UK, this is a natural life moment when this information could be passed to EROs. The advantage of this data it that it would contain all of the necessary personal identifiers that EROs would need to register the citizens to vote. This has already been proposed by the APPG on Democratic Participation (Bite the Ballot, James, and ClearView Research 2016) and supported by the Electoral Commission.26

8.6 This would be a very administratively efficient way of increasing voter registration amongst an under-registered group, who were negatively affected by individual electoral registration. It could potentially save money with voter outreach efforts to them. The success of direct enrolment in Victoria provides good evidence that this would be successful in boosting turnout in the UK. These young citizens could then be provided accompanying civic education lessons in schools. Matteo Bergamini of Shout Out UK told us that this would be an important accompanying reform (albeit beyond the scope of this report):

Political literacy would give us, the next generation, a clear understanding of what politics is, how our society works and why voting is relevant and important.

• Recommendation #13: Citizens are

automatically added to the electoral register when they are issued their National Insurance number ahead of their 16th birthday.

8.7 One concern that has been raised about recommendation 13 is that citizens can be registered as attainers in Scotland and Wales as young as 14. There are separate GDPR requirements for under 16s,27 but these apply only to consent in relation to information society services - effectively social media - rather than to the kind of data that would be dealt with here. The ICO’s guidance on ’Children and the GDPR’ makes it clear

‘As voting is a habitual act, in which voting and non-voting patterns are developed while young, it is very important for

young people to vote at an early age in order to

increase the likelihood that they will continue to vote

throughout their lives.’Taidgh Pledger – National Union for Students

25 Data and information from Freedom of Information Request FOI2020/0011826 https://www.electoralcommission.org.uk/who-we-are-and-what-we-do/our-views-

and-research/our-research/accuracy-and-completeness-electoral-registers/2019-report-2018-electoral-registers-great-britain/completeness-great-britain

27 See: evidence from Dr. Andy Mycock and Toby James to the Scottish Parliament Finance and Constitution Committee, 11 September 2019 http://www.parliament.

scot/parliamentarybusiness/report.aspx?r=12244&mode=pdf p.35-6


28 Freedom of Information Request FOI2020/0016829 https://www.gov.uk/elections-in-the-uk30 Data was obtained from https://www.gov.uk/government/collections/national-insurance-number-allocations-to-adult-overseas-nationals-entering-the-uk, dated accessed 5th February.31 Source: Student Loans Company, Freedom of Information Request 20200120-01. Notes provided by the SLC:

“[1] Study support refers to any application for any financial support, including maintenance and tuition fee loans [2] Postgraduate application figures are derived from accounts created within post-payment system [3] Method of Attendance is used to determine full-time/part-time split for postgraduate information [4] Application type is used to determine full-time/part-time split for undergraduate loans [5] FE support includes Advanced Learner Loan (England), Welsh Government Learning Grant (Wales) and Education Maintanance Allowance (Wales and Northern Ireland) products [6] Study level for change of address customers is determined by repayment customer type as customers could meet multiple criteria of having both undergraduate and postgraduate loans [7] This report covers all domiciles, however application numbers do not cover Scotland - SLC only holds application numbers for England, Wales, Northern Ireland and EU. Address change notifications cover these domiciles plus Scotland. [8] Address update figures are drawn from post-payment system for any address changes to customers’ home address, dates refer to date of address change not date of funding [9] 2019/20 is not a full year - figures are subject to change as the application cycle for this academic year is still open”

that for GDPR purposes ’children’ refers to under 18s: the restrictions on processing other than in relation to information society services apply to all under 18s. GDPR problems should not arise here because the sharing of information would be purposive and limited to that which was required for voter registration. This is another area where limiting the use of the register – in particular to avoid commercial, marketing, profiling and political use – would help to make things clearer.

8.9 There would be a need for EROs in Scotland and Wales to check these new registrations against attainers aged 14 and 15 to avoid duplicate registrations. To administratively simplify the process further, there would be a case for removing attainers from the register in Scotland and Wales aged 14 and 15, provided that a system of AVR could register them in time for the elections.

• Recommendation #14: If citizens are registered automatically on receipt of their NINo, the case for not registering them as attainers prior to this date is considered.

8.10 At present, when citizens are written to ahead of their 16th birthday with their National Insurance number there is no mention of their entitlement to register to vote, even though in parts of the UK they will have been newly enfranchised. Figure 8.1 is the template letter that citizens receive, which was provided by HM Revenue and Customs as a result of a Freedom of Information Request. Neither are there any accompanying leaflets about voter registration.28 In advance of any direct enrolment for citizens, this letter could be altered to notify citizens living in Scotland and Wales how to register to vote.

• Recommendation #15: Letters to citizens in Scotland and Wales notifying them of their National Insurance number ahead of their 16th birthday should include information about voter registration.

Figure 8.1 Template letter received by HMRC

8.11 New National Insurance numbers are also issued to citizens at other points. Should a citizen move to the UK from many other countries then they would be entitled to vote in elections.29 They would often then apply for a National Insurance number in order to work. These applications are processed by the DWP rather than HM Treasury.


8.12 Data from the DWP shows that there were 0.75 million such new numbers issued to adult overseas nationals in the year up to September 2019.30 Irish and qualifying Commonwealth citizens are entitled to vote at all elections and should therefore appear on both Parliamentary and local government registers. EU citizens (except Ireland, Malta and Cyprus) can vote at all other elections and should therefore appear on the local government register. Using government data, we estimate that there would have been 169,354 citizens who could have been added to the parliamentary register in 2019 and 436,263 who would have been added to the local government register. Given that registration rates are 62% for Commonwealth citizens and 54% for EU citizens, a huge proportion of these eligible citizens are likely to be missing from the register. We would estimate that direct enrolment would therefore have increased the completeness of the parliamentary register by 65,000 names and the local government register by 188,000 names. The highest nationality groups of these electors is detailed in Table 8.2.

Potential local government register entries

Potential parliamentary register entries

India 72,330 India 72,330

Italy 51,694 Pakistan 18,486

Poland 42,906 Nigeria 12,669

Spain 37,045 Australia 11,667

Portugal 23,570 Ireland 11,040

Table 8.2: Estimated new entries to the electoral register from new adult NINos, by highest nationality • Recommendation #16: Adult citizens

are automatically added to the electoral register when they apply or are issued a new NINo.

Student Loans Company

8.13 The Student Loans Company (SSLC) also administers a considerable number of transactions from one of the most under-registered groups, students/recent graduates, making it a potentially very useful source. Table 8.3 provides data on the volume of student loan applications, while Table 8.4 provides data on the volume of change of address notifications.31 EROs could therefore be notified when the Student Loans Company receives a new application for study support or a change of address. In order to prevent an overwhelming volume of information being sent to EROs, which most likely will not involve a change of address, the SSLC could instead provide an ‘opt-in’ box for students. Information could then be sent on where students request the SSLC to do so. It could be automatically sent where the SSLC are notified of a change of address.

• Recommendation #17: Citizens should be able to ‘opt-in’ and effectively register to vote when they make a new Student Loan application from the Student Loans Company or make any changes of address.

University registration

8.14 In 2017/18 there were 2.34 million students at UK higher education institutions (Bolton 2019, 14). Although not all of these would be eligible to vote at all elections, measures to enrol students could significantly boost voter registration rates. The All Party Parliamentary Group on Democratic Participation previously recommended that university, further education and higher education students are prompted to register to vote when they annually enrol on their

Academic Year

Full-time undergraduate

support

Full-time postgraduate

support

Part-time undergraduate

support

Part-time postgraduate

support

FE financial support

2015/16 1,489,488 N/A 131,981 N/A 225,500

2016/17 1,487,800 51,351 127,308 18,982 234,297

2017/18 1,529,059 60,947 135,824 35,369 208,124

2018/19 1,531,093 64,790 157,342 37,802 194,714

2019/20 1,487,823 5,161 148,756 20,416 136,818

Table 8.3: Student Loan and Further Education applications

Academic Year

Undergraduate Postgraduate

2015/16 805,976 0

2016/17 928,673 9,145

2017/18 1,043,065 11,025

2018/19 1,112,624 15,370

2019/20 533,046 10,065

Table 8.4: Change of address notifications for student loans


32 https://www.sheffield.ac.uk/news/nr/voter-registration-chris-skidmore-1.665948

33 https://www.researchresearch.com/news/article/?articleId=1374721

course (Bite the Ballot, James, and ClearView Research 2016, 10). Students are required to re-register each year and this provides an excellent opportunity to capture their new address. If voter registration information is built into these forms, then it could be passed on to EROs directly.

8.15 Sheffield City Council Electoral Services worked with Sheffield University to give students the chance to register to vote during their annual university enrolment process. This has been widely flagged as best practice because an estimated 75 per cent and 76 per cent of students were enrolled in the first and second year of the scheme. This compares to rates as low as 13 per cent for other universities.32

8.16 The Higher Education Act 2017 requires universities in England to undertake any steps that the Office for Students (OfS) sets out to help register students. The directions the OfS can give are set by the Minister of State, who acts as a gatekeeper. In February 2018, the Secretary of State for Education set out guidance and the OfS then published the Regulatory Framework for Higher Education in England, which lays out its approach to managing electoral registration in condition E5. A provider will be expected to:

• provide information to an electoral registration officer when asked to do so

• facilitate cooperation and an effective partnership with the electoral registration officer

• provide its students with easily accessible information about how to register to vote

8.17 There is therefore no provision for AVR, but EROs are now provided with information on students that they could use as a local data source to improve the completeness and accuracy of the register. Careful evaluation of the scheme will be required to ascertain whether the scheme is sufficiently directional to lead to widespread adoption of evidence-based best practice. Some flexibility is helpful for universities and electoral administrators, they have argued, because universities and student populations can cut across the jurisdictions of different EROs, especially in London, which makes some activities, such as data sharing, complicated. But with many other pressures facing university

administrations, voter registration is not likely to be given much priority and resources.33 However, there is also clearly further scope to explore registering students by building it into student enrolment. Taidgh Pledger from the National Union of Students told us that:

‘Ever since legislation was first drafted for the move to IER in 2011 NUS has campaigned for the integration of voter registration into student enrolment processes. NUS would favour a mandatory scheme, whereby ERO funds were utilised to ensure that all universities and colleges had the capacity to integrate voter registration with enrolment.’

The National Union of Students also informed us that many universities were not providing EROs with information about their students for them to enrol because they were concerned about data protection issues:

Although several universities have found an approach for mass data sharing with which they felt comfortable, the majority of institutions have not gone down this route because of data protection concerns. This is despite assurances that we received from the Cabinet Office that this need not in any way act as a barrier, and that all data protection issues can be overcome. The government should provide better information to institutions of their responsibilities around data protection so that they can be clear that automatic voter registration schemes would not impinge upon this.

• Recommendation #18: Universities are

stringently monitored to ensure that they are meeting guidance and sanctions are enforced for failures to comply with guidance.

• Recommendation #19: Universities should be required to build voter registration practices into their annual student enrolment processes.

Passports

8.18 The Passport Office is in receipt of a large volume of applications for new passports each year. In 2016, this included 6.49 million applications from UK applicants and 0.44 million international applications for UK


There is further scope to explore registering

students by building it into student enrolment


34 Freedom of Information Request 5733035 Freedom of Information Request FOIR8192.

36 Freedom of Information Request FOI2020/0209737 Freedom of Information Request FOI2020/02097

passports. No breakdown is recorded by the Home Office of the age of the applicants 34 so it is unclear whether these are a notably under-registered group. However, the information taken from the citizen at this moment in time is likely to be complete and accurate. It would therefore be helpful for the citizen to be prompted to register to vote.

8.19 The Home Office also receives a much smaller volume (0.14 million in 2016) of applications for UK citizenship. This is a notable life moment in which a citizen’s eligibility to vote may change. Given the relatively small volume of applications, this information is likely to be useful for EROs and should be made available.

• Recommendation #20: Citizens should be able to ‘opt-in’ and effectively register to vote when they apply for a passport and/or a change in citizenship status.

DVLA

8.20 The DVLA receives many applications for driving licences each year, but other transactions include change of address, driving licence renewals and vehicle tax renewals. Table 8.5 provides data from 2016 to 2019 on the volume of such applications.35 Citizens could be prompted to register to vote at these moments. Provisional driving licence applicants are likely to be younger citizens which would make that data stream highly useful for EROs. There is a legal requirement for citizens to change the address on their driving licence when they move, so this information should be highly reliable. Vehicle tax renewals are much larger in number – but this is a renewal and there is a risk that the volume of transactions would be overwhelming for EROs without thorough piloting. This source would therefore best be discarded.

• Recommendation #21: Citizens should be able to ‘opt-in’ and effectively register to vote when they apply for a full/provisional driving licences and notify the DVLA of a change of address.

Disability benefits

8.21 Personal Independent Payment is provided to citizens who have a health condition or disability. It replaces the Disability Living Allowance.36 Data for the latter is presented in Table 8.6. Meanwhile Attendance Allowance provides for extra costs for citizens with disabilities that may have a physical or mental disability. Data for the latter is presented in Table 8.7.37

8.22 Citizens with disabilities do have lower levels of voter registration, so both application systems would be reasonable opportunities to allow citizens to register to vote.

Year New Claims

2015 92,000

2016 95,000

2017 103,000

2018 113,000

2019 118,000 Table 8.6: Volume of applications for Disability Living

Allowance 2015-2019

Year New Claims

2015 317,000

2016 319,000

2017 307,000

2018 331,000

2019 334,000 Table 8.7: Volume of applications for Attendance Allowance

2015-2019

Provisional drivers licence applications

Full drivers’ licence Drivers licence change of address

Drivers licence renewals

Vehicle tax renewals

2015 Not provided Not provided Not provided Not provided 22,205,890

2016 978,790 876,195 3,982,529 954,720 20,213,054

2017 980,487 950,920 4,059,929 910,702 19,451,661

2018 980,811 880,222 3,859,968 816,068 19,465,661

2019 1,001,002 867,379 4,053,338 928,532 18,384,185

Table 8.5: Volumes of transactions with citizens undertaken by the DVLA


38 Freedom of Information Request FOI2020/0011839 This will replace Housing Benefit, income-related Employment

and Support Allowance (ESA), income-based

• Recommendation #22: Citizens should be able to ‘opt-in’ and effectively register to vote when they submit Personal Independent Payment and Attendance Allowance.

Child Benefit

8.23 Child benefit can be provided to those responsible for bringing up a child under 16, or someone under 20 if they are in approved education or training. New applicants for child benefit will therefore be mostly be parents or guardians of younger children. Table 8.8 shows data on new claimants from 2014/5 to 2018/19. There are usually three-quarters of a million such new applicants.

8.24 This would be a large volume of new

applications for EROs to process, but citizens could be given the opportunity to ‘opt-in’ for their electoral registration details to be updated.

Year New Claims

2014/15 876,663

2015/16 875,126

2016/17 891,383

2017/18 822,353

2018/19 767,401

Table 8.8: Volume of applications for Child Benefit 2015-201938

• Recommendation #23: Citizens should be able to ‘opt-in’ and effectively register to vote when they submit new child benefit claims.

Universal Credit

8.25 One of the major government reforms being rolled out is the move to Universal Credit, which has been phased in and is replacing many other state benefits.39 Data from the DWP in Table 8.9 below shows how there has been a considerable rise in applications for the scheme so that 2.5 million people were on Universal Credit by October 2019, with over 400,000 of those being in the 16-24 age range. Citizens could therefore be offered the opportunity to register to vote when they make new applications for Universal Credit. This information would help to boost registration rates amongst those in lower socio-economic groups, who are also more likely to be private renters.

• Recommendation #24: Citizens should be able to ‘opt-in’ and effectively register to vote when they submit new claims for Universal Credit.

Other DWP Transactions

8.26 Some transactions are few in number but have a symbolic importance. The Life in the UK Test is a computer-based test that is one of the requirements for anyone seeking Indefinite Leave to Remain in the UK or naturalisation as a British citizen. The number of transactions is relatively small at 126,021 in 2016/17, but given that this is a moment when a citizen’s status on the franchise

Age band 16-24 25-49 50 Plus Unknown/Missing Total

October 2018 263,830 769,367 252,363 565 1,286,127

November 2018 281,343 832,555 274,656 634 1,389,187

December 2018 306,365 921,347 306,720 710 1,535,143

January 2019 317,629 974,286 325,365 764 1,618,043

February 2019 327,651 1,032,817 345,311 830 1,706,607

March 2019 343,434 1,106,116 365,902 830 1,816,280

April 2019 361,581 1,189,023 389,415 880 1,940,904

May 2019 370,636 1,248,736 405,930 948 2,026,245

June 2019 385,478 1,336,282 429,194 58 2,151,015

July 2019 399,607 1,402,462 446,932 65 2,249,063

August 2019 415,963 1,469,051 465,298 55 2,35,0373

September 2019 428,311 1,540,792 485,482 64 2,454,646

October 2019 (r) 439,146 1,604,165 503,610 54 2,546,973

Table 8.9 Volume of Universal Credit applications from October 2018-19

40 Freedom of Information Request FOI2020/0011852 | Finding the ‘Life Moments’: New National Data Sources for Direct Enrolment

might change, it would be an appropriate opportunity to register them to vote. An application for British citizenship would be another opportune moment with a similar volume of applications.

• Recommendation #25: Citizens should be able to ‘opt-in’ and effectively register to vote when they submit claims for Indefinite Leave to Remain.

Tax Self Assessments

8.27 A citizen must return a tax self-assessment to HMRC if they have been self-employed, a partner in a business or in receipt of additional income sources. Information from a FOI in Table 8.10 shows that reached 12.5 million in 2018/19.40 This is another potential

‘life moment’ where citizens will also be providing all of the information that they would need to register to vote.

8.28 The volume of applications, however, is very high relative to number of likely registrants. These are not likely to be groups of under-registered citizens. EROs could therefore be overwhelmed with the volume of information provided. It might still be advantageous to have an ‘opt-in’ button in which citizens could selectively have their information passed to EROs if they thought that they were likely to be incorrectly registered.

Tax Year Submissions

2014/15 11,465,000

2015/16 11,870,000

2016/17 12,177,000

2017/18 12,464,000

2018/19 12,536,000

Table 8.10 Volume of Tax Self Assessments 2014/2015-2018/2019

• Recommendation #26: Citizens should be able to ‘opt-in’ and effectively register to vote when they submit their tax return.

School Information

8.29 The Automatic Electoral Registration (School Students) Bill was introduced into the Lords in June 2016 with the proposal that all students in schools would be registered automatically. The Bill specified that all EROs would have a responsibility to ensure that students aged 16 or above would be on the electoral register. The Secretary of State would be given a power to require schools to return information to EROs. Students would only be registered if they had the consent of the parents and schools would be required to confirm whether this was the case to EROs. The Bill did not reach second reading before Parliament was prorogued, but provides a helpful example of how automatic voter registration could work in schools.

8.30 There remains considerable value in this approach. However, if voter registration is made automatic for 16-year olds ahead of the issuing of their NINo then it would not be necessary. Efforts could instead focus on citizenship lessons


and students being reminded to change their address if they have moved. However, if direct enrolment at the point of issuing a NINo is not followed then this would be a good second choice option to pursue.

Devolved Administrations

8.31 A number of public services and bodies are under the control of the Welsh government and Scottish government. Transactions that citizens have with these services could therefore also be made available to EROs. The advantage of these sources is that they could be used by the devolved governments even if the UK government does not wish to pursue this approach across the whole of the UK. A consideration of these possible sources was beyond the scope of this report.

8.32 It is likely that the Welsh or Scottish specific

data is not more complete or accurate than the existing data already reviewed, however. It would be advantageous for EROs working in Scotland and Wales, where AVR may be introduced for local government registers, that will also be used in the devolved parliamentary elections, to have access to these national data sources.

• Recommendation #27: National data sources, such as those identified in recommendations 13-26 should be made available to EROs in Scotland and Wales, even if they not made available UK-wide.

Private sector sources

8.33 Private sector companies maintain a vast range of datasets that could be potentially useful for EROs. Utility companies, for example, would be amongst the first to be notified when citizens move house because they will need to access their broadband, water or electricity at the new house. Meanwhile credit reference agencies buy all electoral registers, including the full electoral register, thereby being better able to spot double registrations.

8.34 Some respondents thought that it would ‘not be necessary’, however, to use this information for AVR if other public sector data sources were made available to EROs.

It is also likely that citizens may feel that their privacy is breached if information about their private purchases are sent to governmental bodies. The possibility of misuse of this information and the related profiling information, in relation for example to benefits entitlement, could be significant.

8.35 There could be a statutory requirement for credit reference agencies to provide a combined version of all electoral registers, which the Electoral Commission could make available to EROs. This would be helpful to provide an overall analysis of the electoral register. There might then be scope for passing information onto EROs about duplicate registrations and possible missing electors.

• Recommendation #28: Credit reference agencies should be required to provide a copy of the combined electoral register to the Electoral Commission for the purposes of assessing and increasing the accuracy and completeness of the register.

SummaryIn summary, AVR in the UK would require selectively choosing appropriate data sources according to criteria such as reliability, completeness and traffic volume. Using new transactions with governmental organisations rather than data-mining to compare databases would be the most efficient and effective way forward, as well as reducing some of the related privacy risks. The automatic enrolment of citizens when they are issued NINos ahead of their 16th birthday could address the major under-registration issues with young people across the UK. Providing opt-in opportunities to register to vote at other government services will also help EROs to maintain the electoral register to improve overall completeness and accuracy, without them being overwhelmed with data.

1 sub

Overall, this report makes 28 recommendations to modernise the UK’s electoral machinery for an inclusive twenty-first century democracy.

Conclusions and Recommendations | 55

This final chapter summaries the conclusions and recommendations.

9.1 The electoral registration system used across the UK was designed over a century ago, in a context with specific needs and challenges. There has been much bolt-on modernisation since the turn of the century with continuous registration, individual electoral registration and canvass reform being introduced at different times across the UK. There are major concerns about whether the current system is fit for purpose, however. This report has identified nine problems with the status quo:

1. Missing Millions: there are millions of eligible citizens who are incorrectly registered or have their names missing entirely from the electoral register.

2. Marginalised Groups: there is considerable inequality in voter registration rates.

3. Voters turned away: citizens are frequently unable to vote because of registration issues.

4. Funding crises: electoral services are needing to compromise on service quality for financial reasons.

5. Last-minute pressures: there are major surges in voter registration applications close to election day, which are generating a risk on the system.

6. Duplicate applications: citizens are attempting to register multiple times because of confusion.

7. Use of the register by credit reference agencies: credit reference agencies are paying disproportionately low fees for rich, publicly produced data about the process and their status.

8. The use of the edited register: information from the register is used in an unrestricted way, potentially violating the privacy of citizens.

9. Anonymous registration: there remains the need to protect vulnerable individuals from the publication of their data.

9.2 This report shows that AVR is the norm internationally in most leading democracies. Many Anglosphere democracies, who have historically not had AVR, partly because of their colonial legacy with the UK from which their system was partially derived, have already embarked on using direct enrolment for specific groups of under-registered voters.

Ensuring privacy

9.3 To make direct enrolment possible, and close existing loopholes with the integrity of the system, the report has recommended that:

• Recommendation #1: The fee that credit reference agencies pay for access to the full register should be increased in line with inflation since 2001. The case for a higher fee to fund electoral registration modernisation should be considered.

• Recommendation #2: The sale of electoral

register data by credit reference agencies to third parties should be strictly prohibited.

• Recommendation #3: The Open/Edited electoral register should be abolished.

• Recommendation #4: The range of attesters for anonymous registration should be expanded to reflect victims’ lived experiences of abuse.

• Recommendation #5: Anonymous registration should be expanded to at least five years.

9 | Conclusions and Recommendations

56 | Conclusions and Recommendations

• Recommendation #7: If automatic voter registration is introduced, then directly enrolled citizens must not have their details added to the open/edited register (should that register not be abolished).

• Recommendation #8: Legal restrictions on the use of the electoral register should be considered should GDPR laws change following Brexit.

Philosophical issues and organisational issues

9.4 The question of whether AVR should be introduced is a philosophical one. Concerns have been raised by those consider it voter registration should be an individual not state responsibility. Democratic theorists argue that democracy requires a level playing field on election day. It is important to ensure that there is a full and accurate electoral register so that some groups do not suffer from disproportionally low levels of voter registration. It is therefore recommended that:

• Recommendation #6: Voter registration should remain compulsory across the UK

• Recommendation #9: Local EROs should continue to be the main custodians of the electoral register in the geographical area, with support from the Electoral Commission.

Methods for introducing AVR

9.5 There are a variety of options for introducing AVR. One would be the introduction of a compulsory population register, but this reform would involve much wider issues than electoral registration so was not recommended. Alternatives would be:

• Recommendation #10: The establishment of a central electoral register based on the DWP CIS is explored in the long-term.

9.6 More immediate reforms could include:

• Recommendation #11: Direct enrolment for specific under-registered groups when they undertake new transactions with other government services

• Recommendation #12: Citizens are prompted to register to vote when they access certain government services.

9.7 ‘Data matching’ – in which EROs compare the register to new national sources - is not likely to be a cost-efficient way forward, given the pressures that public services are under. However, there should be targeted interventions to register the most under-registered groups which will have minimal impact on wider public services.

Conclusions and Recommendations | 57

Direct enrolment to increase accuracy and completeness

9.8 Direct enrolment could be pursued following the completion of the transition to the reformed canvass. The report has therefore recommended:

• Recommendation #13: Citizens are automatically added to the electoral register when they are issued their NINo ahead of their 16th birthday.

• Recommendation #14: If citizens are registered automatically receipt of their NINo, the case for not registering them as attainers prior to this date is considered.

• Recommendation #15: Letters to citizens in Scotland and Wales notifying them of their National Insurance number ahead of their 16th birthday should include information about voter registration.

• Recommendation #16: Adult citizens are automatically added to the electoral register when they apply or are issued a new NINo.

9.9 All new enrolments should be notified that they have had their name added to the electoral register at least four weeks before they are added. This will give them time to register anonymously should they need to. They should be excluded from the open/edited register.

‘Opt-in’ voter registration at other services

9.10 Citizens should be provided with an ‘opt-in’ opportunity to directly enrol when they access specific other services. The report has therefore recommended that:

• Recommendation #17: Citizens should be able to ‘opt-in’ and effectively register to vote when they apply for a new Student Loan application from the Student Loans Company or make any changes of address.

• Recommendation #18: Universities are stringently monitored to ensure that they are meeting guidance and sanctions are enforced for failures to comply with guidance.

• Recommendation #19: Universities should be required to build voter registration practices into their annual student enrolment processes.

• Recommendation #20: Citizens should be able to ‘opt-in’ and effectively register to vote when they apply for a passport and or a change in citizenship status.

• Recommendation #21: Citizens should be able to ‘opt-in’ and effectively register to vote when they apply for a full/provisional driving licences and notify the DVLA of a change of address.

• Recommendation #22: Citizens should be able to ‘opt-in’ and effectively register to vote when they submit Disability Living Allowance and Attendance Allowance.

• Recommendation #23: Citizens should be able to ‘opt-in’ and effectively register to vote when they submit new child benefit claims.

• Recommendation #24: Citizens should be able to ‘opt-in’ and effectively register to vote when they submit new claims for universal credit.

• Recommendation #25: Citizens should be able to ‘opt-in’ and effectively register to vote when they submit claims for Indefinite Leave to Remain.

• Recommendation #26: Citizens should be able to ‘opt-in’ and effectively register to vote when they submit their tax return.

• Recommendation #27: National data sources, such as those identified in recommendations 12-26 should be made available to EROs in Scotland and Wales, even if they not made available UK-wide.

• Recommendation #28: Credit reference agencies should be required to provide a copy of the combined electoral register to the Electoral Commission for the purposes of assessing and increasing the accuracy and completeness of the register.

58 | Appendix A

• What are the problems with the existing system of electoral registration?

• What alternative systems have been used overseas? How effective have they proven?

• What sources of data could be used for automatic voter registration? For example, information from other public bodies? Information from the private sector?

• Would you have any concerns about cyber-security? If so, how could they be redressed?

• What concerns would you have, if any, about data privacy? If so, how could it be redressed?

• How should specific privacy concerns, for example, personal data of victims of violence and domestic abuse, be addressed?

• What concerns, if any, do you have about the use of the electoral register? For example, by government departments or the sale of electoral register? How can they be mitigated?

• Should voter registration be opt-in or opt-out? To what extent can consent and active citizen choice be achieved?

• Any other comments.

Appendix A: Survey Questions

Appendix B | 59

The following individuals or organisations provided input into the report for which we are very grateful:

• Areeq Chowdhury, WebRoots Democracy • Pascal Crowe, Open Rights Group • Lucy Hadley, Women’s Aid • Jess Garland, Electoral Reform Society • Peter Stanyon, Association for Electoral Administrators • Grant Hazell, Democracy Counts • Dr Michael A Ward • Luke Butcher, National Election Commission of the Republic of Korea • Charlie Fisher, Teignbridge District Council • Anna Last, Ipswich Borough Council • Lisa Vines, Lancaster City Council • Nicky Gilligan, East Staffordshire Borough Council • Steve Daynes, Braintree District Council • David Levine, Alliance for Securing Democracy • Stuart Gurthie, Norwich City Council • Thomas Henry, Norwich City Council • Jean Harrison, individual citizen • Tobias William Reynolds, University of Salford (student) • Peter Clark • Stella-Maria Thomas • Ann Taylor, Merthyr Tydfil County Borough Council • Matteo Bergamini, Shout Out UK • Ben Graham Jones, International Election Analyst and Observer • Dr Ian Graham, University of Edinburgh • Sue Rodden, Broxtowe Borough Council • Taidgh Pledger, National Union of Students • Nathaniel Reader, Victorian Electoral Commission

Appendix B: Respondents

41 https://www.gov.uk/performance/services42 https://www.gov.uk/performance/services60 | Appendix C

Life moment Organisation Annual volume of applications

Application for full-time study support Department for Education / Student Loans Company 1.92 million

Application for part-time study support Department for Education .11 million

Student loans: voluntary repayments Department for Education / Student Loans Company .65 million

Applications for Further Education (FE) financial support

Department for Education / Student Loans Company .07 million

Drivers licence change of address DVLA 3.03 million

Provisional driving licence DVLA 1.12 million

Drivers licence changes DVLA 3.0 million

New Pension Credit Claims Dept for Work and Pensions 0.10 million

Disability Living Allowance: new claims Dept for Work and Pensions .18 million

State pension: new claims Dept for Work and Pensions .43 million

Income support: new claims Dept for Work and Pensions 0.24

Budgeting Loans Dept for Work and Pensions 1.43 million

Carer’s allowance: new claims Dept for Work and Pensions .28 million

Attendance allowance: new claims Dept for Work and Pensions .35 million

Jobseeker’s Allowance: new claims Dept for Work and Pensions 1.01 million

Child maintenance transactions Dept for Work and Pensions No data

Child benefit transactions HM Revenue & Customs 1.17 million

Self assessment tax submissions HM Revenue & Customs 7.04 million

Domestic applications for UK passports Home Office 6.49 million

International applications for UK passports Home Office .44 million

Home Office travel documents for non-British citizens

Home Office .35 million

Life in the UK Test Home Office .13

Applications for British citizenship Home Office .14

Table A1: Select new transactions with citizens, based on data for 2016-1741

Vehicle tax renewals DVLA 47.1 million

Drivers licence renewals DVLA 4.1 million

Existing Pension Credit Claims Dept for Work and Pensions 1.84 million

Disability Living Allowance: existing claims Dept for Work and Pensions 2.4 million

State pension: existing claims Dept for Work and Pensions 12.9 million

Carer’s Allowance: existing claims Dept for Work and Pensions 1.22 million

Attendance Allowance: existing claims Dept for Work and Pensions 1.59 million

Income Support: existing claims Dept for Work and Pensions .60 million

Table A2: Select renewal transactions, based on data for 2016-1742

Appendix C: Government Transactions with Citizens

Bibliography | 61

192.com. 2020. https://www.192.com/misc/privacy-policy/.

AEC. 2019. Direct Enrolment and Update. https://www.aec.gov.au/Enrolling_to_vote/About_Electoral_Roll/direct.htm.

Benett, Asa. 2013. “Alastair Crawford On How He Built £6.5m-A-Year Directory Website 192.Com.” Huffington Post. https://www.huffingtonpost.co.uk/2013/11/22/192-alastair-crawford_n_4324020.html.

Bite the Ballot, Toby S. James, and ClearView Research. 2016. Getting the Missing Millions Back on the Electoral Register: A vision for voter registration reform in the UK. London: All Parliamentary Party Group on Democratic Participation.

Black, Jerome H. 2003. From Enumeration to the National Register of Electors: An Account and an Evaluation. Vol. 9, Choices. Quebec: Institute for Research on Public Policy.

Bolton, Paul. 2019. Higher education student numbers. London: House of Commons Library.

Brennan Center. 2019. Automatic Voter Registration: A Summary: Brennan Center for Justice.

Brown, Mitchell , Kathleen Hale, Robert J. Smith, and Lindsey Forson. 2020, forthcoming. “Capacity to Address Natural and Man-made Vulnerabilities: The Administrative Structure of U.S. Election System Security.” Election Law Journal, p.

Carr, Simon. 2016. “Directory giant 192 slammed for ‘stalker’ app that tracks down former lovers.” The Mirror, 17 February 2016. https://www.mirror.co.uk/news/uk-news/directory-giant-192-slammed-stalker-7384702.

Center for Internet Security. 2018. A Handbook for Elections Infrastructure Security. New York: CIS.

Clark, Alistair. 2014. “Investing in Electoral Integrity.” In Advancing Electoral Integrity, edited by Pippa Norris, Richard Frank and Ferran Matinez I Coma, 165-88. New York: Oxford University Press.

———. 2015. “Public Administration and the Integrity of the Electoral Process in British Elections.” Public Administration, 93, (1), p. 86-102.

Clark, Alistair, and Toby S. James. 2017. “Poll Workers.” In Election Watchdogs, edited by Pippa Norris and Alessandro Nai. New York: Oxford University Press.

Coppedge, Michael, John Gerring, Carl Henrik Knutsen, Staffan I. Lindberg, Jan Teorell, David Altman, Michael Bernhard, M. Steven Fish, Adam Glynn, Allen Hicken, Anna Lührmann, Kyle L. Marquardt, Kelly McMann, Pamela Paxton, Daniel Pemstein, Brigitte Seim, Rachel Sigman, Svend-Erik Skaaning, Jeffrey Staton, Steven Wilson, Agnes Cornell, Lisa Gastaldi, Haakon Gjerløw, Nina Ilchenko, Joshua Krusell, Laura Maxwell, Valeriya Mechkova, Juraj Medzihorsky, Josefine Pernes, Johannes von Römer, Natalia Stepanova, Aksel Sundström, Eitan Tzelgov, Yi-ting Wang, Tore Wig, and Daniel Ziblatt. 2019. V-Dem [Country-Year/Country-Date] Dataset v9.

Dahl, R. 1971. Polyarchy: Participation and Opposition. New Haven: Yale University Press.

Deputy Prime Minister. 2011. Individual Electoral Registration. London: The Stationary Office.

DWP. 2018. Customer Information System (CIS): the information held about you. London: DWP.

Electoral Commission. 2013. Data mining pilot – evaluation report. London: Electoral Commission.

———. 2015. The May 2015 UK elections: Report on the administration of the 7 May 2015 elections. London: Electoral Commission.

———. 2016. Performance standards for Electoral Registration Officer. London: Electoral Commission.

Bibliography

62 | Bibliography

———. 2017a. Analysis of the December 2016 electoral registers in the United Kingdom.

———. 2017b. Electoral registration at the June 2017 UK general election. London: Electoral Commission.

———. 2018. Electoral registration in Great Britain: 2017 and 2018 annual canvasses. London: Electoral Commission.

———. 2019a. Accuracy and completeness of the 2018 electoral registers in Great Britain. London: Electoral Commission.

———. 2019b. Guidance for Electoral Registration Officers: Part 2 – The registration framework.

———. 2019c. Guidance for Electoral Registration Officers: Part 4 – Maintaining the register throughout the year. London: Electoral Commission.

———. 2019d. Modernising electoral registration in the UK: feasibility studies. London: Electoral Commission.

———. 2019e. Supply of the Electoral Register. Electoral Commision, https://www.electoralcommission.org.uk/sites/default/files/2019-11/Supply%20of%20the%20Electoral%20Register.pdf.

Ernst and Young. 2019. Department of Motor Vehicles ─ Independent System Assessment Business Process Assessment Report. Sacramento, CA: Department of Finance.

Ferguson, Kate. 2018. “Reviving plans for ID cards would prevent repeat of Windrush scandal, say former Home Secretaries.” Daily Mail, 30 April 2018. https://www.dailymail.co.uk/news/article-5673421/Charles-Clark-Alan-Johnson-urge-introduction-ID-cards-Windrush.html.

Franklin, Daniel P., and Eric E. Grier. 1997. “Effects of motor vehicles legislation: Voter turnout, registration, and partisan advantage in the 1992 presidential election.” American Politics Quarterly, 25, (1), p. 104-17.

Goodhart, David, and Richard Norris. 2018. The UK Border Audit. London: Policy Exchange.

Hansard. 2016. Automatic Registration: UK Elections, 29 June 2016. London: Houses of Parliament.

Hencke, David. 2011. “Privacy? It’s a dangerous myth.” British Journalism Review, 22, (3), p. 43-8.

HM Government. 2013. Improving the completeness & accuracy of the electoral register through data matching: An evaluation of pilots testing data matching for the purposes of identifying new electors London: Cabinet Office.

Hone, Alexander. 2014. Electoral registration: Penalties for providing false information or refusing to provide information. London: House of Commons Library.

James, Toby S. 2012. Elite Statecraft and Election Administration: Bending the Rules of the Game. Basingstoke: Palgrave Macmillan.

———. 2013. “Fixing failures of UK electoral management.” Electoral Studies, 32, (4), p. 597-608.

———. 2014. “The Spill-Over and Displacement Effects of Implementing Election Administration Reforms: Introducing Individual Electoral Registration in Britain.” Parliamentary Affairs, p. 281.

———. 2017. “The Effects of Centralising Electoral Management Board Design.” Policy Studies, 38, (2), p. 130-48.

———. 2019. “Better Workers, Better Elections? The Workforces of Electoral Management Bodies and Electoral Integrity.” International Political Science Review, 40, (3), p. 370–90.

———. 2020. Comparative Electoral Management: Performance, Networks and Instruments. London and New York: Routledge.

James, Toby S., and Alistair Clark. 2019. “Electoral integrity, voter fraud and voter ID in polling stations: lessons from English local elections.” Policy Studies, p. 1-20.

———. 2020. “Delivering electoral integrity under pressure: local government, electoral administration, and the 2016 Brexit referendum.” Local Government Studies, p.

James, Toby S., and Holly Ann Garnett. 2020. “Introduction: the case for inclusive voting practices.” Policy Studies, 41, (2-3), p. 113-30.

James, Toby S., and Tyrone Jervier. 2017. “The cost of elections: the effects of public sector austerity on electoral integrity and voter engagement.” Public Money & Management, 37, (7), p. 461-8.

JSCEM. 2012. Administration of the 2011 NSW Election and Related Matters. Sydney, NSW: NSW Parliament.

Bibliography | 63

Labour Party. 2005. Britain forward not back. London: Labour Party.

May, Theresa. 2010. Identity Documents Bill, second reading. London: Hansard.

Morris, Kevin, and Peter Dunphy. 2019. AVR Impact on State Voter Registration: Brannan Center.

Nickerson, David W. 2008. “Is voting contagious? Evidence from two field experiments.” American Political Science Review, 102, (1), p. 49-57.

O’Leary, Cornelius. 1962. Elimination Of Corrupt Practices In British Elections,1868-1911 London: Clarendon Press.

Rakich, Nathaniel. 2019. What Happened When 2.2 Million People Were Automatically Registered To Vote. FiveThirtyEight, https://fivethirtyeight.com/features/what-happened-when-2-2-million-people-were-automatically-registered-to-vote/.

Reader, Nathaniel. 2015. “Assessing the policy impact of Australia’s parliamentary electoral matters committees: a case study of the Victorian Electoral Matters Committee and the introduction of direct electoral enrolment.” Parliamentary Affairs, 68, (3), p. 494-513.

Rosenblatt, G., P. Thompson, and D. Tiberti. 2012. “The Quality of the Electoral Registers in Great Britain and the Future of Electoral Registration.” Parliamentary Affairs, 65, (4), p. 861-76.

Rosenstone, Steven J., and Raymond E. Wolfinger. 1978. “The Effect of Registration Laws on Turnout.” American Political Science Review, 72, (1), p. 22-45.

Select Committe on Political and Constitutional Reform. 2014. Voter engagement in the UK. London: The Stationery Office.

Seymour, Charles. 1915. Electoral Reform in England and Wales: David & Charles Reprints.

Sheppard, Jill. 2015. “Compulsory voting and political knowledge: Testing a ‘compelled engagement’ hypothesis.” Electoral Studies, 40, p. 300-7.

The Public Administration and Constitutional Affairs Committee. 2017. Lessons Learned from the EU Referendum. Translated by Public Administration and Constitutional Affairs Committee. London, United Kingdom: Parliament of the United Kingdom.

The Week. 2018. “The pros and cons of ID cards for UK citizens.” p.

Thomas, Richard, and Mark Walport. 2008. Data Sharing Review Report.

UK Government. 2011. £11 million saved as electors database plan abandoned. UK Government, https://www.gov.uk/government/news/11-million-saved-as-electors-database-plan-abandoned.

United States Senate Intelligence Committee. 2019a. “Russia’s Use of Social Media.” In, edited by United States Senate Intelligence Committee.

———. 2019b. “Russian Efforts Against Election Infrastructure.” In, edited by United States Senate Intelligence Committee.

Vasilogambros, Matt. 2019. “Glitches in California Embolden Automatic Voter Registration Foes.” Pew, 17 October 2019. https://www.pewtrusts.org/en/research-and-analysis/blogs/stateline/2019/10/17/glitches-in-california-embolden-automatic-voter-registration-foes.

Victorian Electoral Commission. 2017. Direct Enrollment Report. Melbourne: Victorian Electoral Commission,.

White, Geoff. 2012. “Identity theft fraud gang jailed.” Channel 4 News, 8 June 2012.

White, Isobel, and Alexander Horne. 2014. Supply and sale of the electoral register. London: House of Commons Library.

Wilks-Heeg, Stuart. 2012. A literature review for the Cabinet Office Electoral Registration Transformation Programme. London: Cabinet Office.

Wolfinger, Raymond E., and Jonathan Hoffman. 2001. “Registering and Voting with Motor Voter.” PS: Political Science and Politics, 34, (1), p. 85-92.

Wolfinger, Raymond E., and Steven J. Rosenstone. 1980. Who Votes? New Haven, CT: Yale University Press.

Zuboff, Shoshana. 2019. The age of surveillance capitalism: The fight for a human future at the new frontier of power. London: Profile Books.

Design by

www.penandmousedesign.co.uk